Podchaser Logo
Home
Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Released Wednesday, 23rd November 2022
 1 person rated this episode
Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Wi-Peep - FBI purchased Pegasus, Passkey support directory, Quantum decryption deadline, Firefox 107

Wednesday, 23rd November 2022
 1 person rated this episode
Rate Episode

Episode Transcript

Transcripts are displayed as originally observed. Some content, including advertisements may have changed.

Use Ctrl + F to search

0:00

It's time for security now. Steve Gibson

0:02

is here with lots to talk about.

0:05

Red Hat cryptographically signing

0:07

its zips. do you do such a thing?

0:10

We'll talk about the FBI. Apparently, they

0:12

tried to use Pegasus.

0:15

How legal is that? And then we're

0:17

gonna talk about why people, a new way to

0:19

map, WiFi access points

0:21

or more threateningly to attract people

0:24

using WiFi devices all that

0:26

more coming up next on security now.

0:30

Podcasts you love. From

0:32

people you trust. Thanks.

0:35

is twitch. This

0:41

is security now with Steve Gibson, episode

0:44

eight hundred ninety eight. recorded

0:46

Tuesday, November twenty second twenty

0:48

twenty two. WiPeep.

0:52

Security now is brought to you by Tinuum.

0:54

Tanium unites operations and

0:57

security teams with a single platform

0:59

that identifies where all your IT data

1:01

is patches every device you

1:03

own in seconds and implements critical

1:06

security controls all from a single

1:08

pane of glass. Are you ready to protect

1:10

your organization from cyber threats?

1:12

Learn more at tennium dot

1:14

com slash Twitter. And

1:16

by Barracuda. Barracuda

1:19

has identified thirteen types of

1:21

email threats and how cyber criminals

1:24

use them every day. fishing, conversation

1:26

hacking, ransomware. Plus ten more

1:28

tricks cyber criminals use to steal money

1:31

from your company or personal information

1:33

from your employees and customers. get

1:35

your free ebook at barracuda dot

1:37

com slash security now.

1:40

And bye. SecureWorks.

1:43

Are you ready for the inevitable cyber

1:45

threats? SecureWorks detects

1:48

evolving adversaries and defends against

1:50

them? with a combination of security

1:52

analytics and threat intelligence directly

1:55

from their own counter threat unit.

1:57

Visit secure work dot com slash to

1:59

get a free trial of tajus extended

2:02

detection and response, also known

2:04

as XDR. It's

2:06

time for security. Now, the show

2:08

we cover your security, your privacy, your

2:11

online exploits, your

2:13

offline deploy TWiT

2:16

this guy right here, mister Steve Gibson.

2:18

Hi, Steve. Leo agreed to

2:20

be with you again. That's always good

2:22

to see. What is this? The pre Thanksgiving episode?

2:25

It is. Of yeah. Yeah.

2:27

Now And we're almost in

2:29

the nine hundreds, which is a scary

2:31

place to be. Actually, it was

2:33

interesting because when I got

2:36

Elaine's transcript last week, she

2:38

said this was eight hundred

2:40

and ninety seven. Yeah. and

2:42

she she reminded me. She said,

2:44

okay. That means that

2:46

we are one hundred and two episodes from

2:49

nine ninety nine. Mhmm. And

2:51

there are fifty one episodes per

2:54

year -- Mhmm. -- because we we skip

2:56

one for the holidays. She's paying attention,

2:58

which means exact precisely,

3:02

two more years. Wow. I have security

3:04

now. I'll put that in my calendar.

3:07

So, oh, honey, I'm not gonna let you forget.

3:10

So, by then, you might say,

3:12

oh, I'd like to keep doing this. You

3:14

know, Leo, I may have the hang of it

3:16

by then. And so if you like,

3:19

not not that big a deal. Well, as somebody

3:21

who just quit the radio show, after

3:24

nineteen years of doing that, I

3:26

can kinda understand after

3:28

a while you get to a point where it's like, you know, I've

3:30

done everything I'm gonna do. and imagine.

3:32

Now I would say you could sleep in on except

3:35

that the show didn't start till eleven. So if

3:37

this really changes your your

3:39

sleeping habits, then we have a different problem. I

3:41

get to do stuff on Saturday, which is I

3:44

mean, I've worked -- Yeah. -- weekends

3:46

for nineteen years. That's a

3:48

long time. Yes. In fact,

3:50

what was happening was you were only

3:52

working six hours odd,

3:55

you know, two days, Saturday and Sunday

3:57

for three hours. Yeah. And then you

3:59

and I were meeting once a

4:01

month up in Toronto. Yeah. You

4:03

you were spending four days up there to

4:05

record for I mean, I'm getting PTSD.

4:07

Just hearing about it. This is -- And you

4:10

-- crazy. And you had three

4:12

unfilled week day

4:14

weeks And so you said, you

4:17

know, let's do some shows. I got a

4:19

lot of time on my hands here. What's I

4:21

thinking? Let's well, aren't you glad

4:23

now? that you have a

4:25

podcast -- I am. -- work. I was I've been

4:27

telling people this is the first time I I've

4:30

I've I've not been working for anybody in

4:32

my whole working life. I'm

4:35

working for myself for the first time ever,

4:37

something you know a lot about, but well, except

4:39

now you have a wife. So Well, as

4:41

Patrick Norton once told me, because I said, I don't

4:43

wanna work for the man. He said Leo, there's always a

4:45

man. And in this case,

4:47

the man is a woman, but still now.

4:49

She we're partners. But it but it is

4:51

kind of interesting that I've

4:54

been a, you know, employee, a ten

4:56

ninety or AW2 employee. for

4:58

my since I was sixteen years old.

5:00

So that's so that is a big change.

5:03

I think the podcast thing might work out. That's

5:05

all I'm saying. It might turn

5:07

out to be something. I don't need to keep this job

5:09

anymore. Yeah. might. So

5:12

we're gonna note this week

5:14

many things. we've got a new version

5:16

of Firefox. Google

5:18

recently reached a nearly four hundred

5:20

million dollar user

5:22

tracking settlement Steve got

5:24

some interesting legislative things to talk about

5:27

during these next couple hours. Red Hat

5:29

has started cryptographically signing

5:31

its ZIP distributions Like,

5:33

what? You can sign a zip? Well, not

5:35

really. But the FBI purchased

5:38

turns out the nefarious Pegas

5:40

software or the spyware,

5:43

think TWiT just to kinda

5:45

see what it's about. Uh-huh. Greece

5:47

paid seven million euros

5:50

for a similar spire

5:52

called Predator, pass

5:54

keys has a directory listing

5:56

the sites where they can be used. So

5:58

That will be exciting. The OMB,

6:01

the US Office of Management and Budget,

6:03

has decreed a

6:05

quantum decryption deadline. Oh.

6:08

Leo, yes. Of course, we're all gonna

6:10

pay attention to that. Leo, thirty

6:12

three, speaking of paying attention

6:15

to the FTC, thirty three US

6:17

Steve attorneys general have asked the

6:20

FTC to get serious, my

6:22

friends, about online privacy

6:25

regulation We'll see how that turns

6:27

out. We've got some engaging listener

6:29

feedback, and spin ride is

6:31

finally a day

6:33

or two, away from its

6:36

final testing. What's in?

6:38

Yeah. It's done. It's I I'll explain.

6:41

I'll explain. I've got I have a couple of

6:43

drives here. There's three three drives

6:45

which are weird

6:47

and you should not write to these drives.

6:49

So I'll I'll explain about that. Oh.

6:54

And then we're gonna wrap up by examining

6:56

some chilling research, which

6:59

allows the physical location

7:02

in three d space

7:04

of every WiFi device

7:08

within its range,

7:10

like within a multi Steve

7:12

building or whatever

7:14

facility to be accurately located

7:18

within a meter or so by

7:20

someone simply walking past

7:22

or flying a tiny drone.

7:25

for about twenty bucks. So

7:27

that's the white peep thing. So

7:30

we're gonna talk about all this, and we have

7:32

a picture of the week that had you almost

7:34

falling off your chair. It was pretty funny. It

7:36

was it's a pretty It was kids. I

7:39

liked it. I think a good podcast.

7:41

All coming up on this fine

7:44

eight hundred and ninety eight edition

7:46

of security. Now that's kind of

7:48

amazing. Isn't it? Are you at

7:51

ninety eight? Well,

7:54

you know, it's funny because the the last

7:56

tech I show is December eighteenth.

7:58

And it's gonna be, I think, episode nineteen fifty

8:01

five. I'm one shy of my birth year. I thought

8:03

if I could just do or actually one it'll be nineteen

8:05

fifty four and then the best that will be nineteen fifty

8:07

five. but just do one more.

8:09

That's okay. That's okay. Our

8:11

show today brought to you by Tanium.

8:14

Love these guys. They're

8:16

Their position about all this security

8:19

stuff is that get ready for this.

8:21

The industry's approach to cybersecurity

8:24

has a fundamental flaw.

8:27

And I think you'll agree when I say

8:29

what it is, IT management security

8:32

point tools really don't do

8:34

it all. They only offer a small piece of

8:36

the solution needed to protect your environment.

8:38

Many of them promise they

8:40

can stop all breaches. They just

8:42

can't. The

8:44

key to a we've always talked about

8:46

layered security. Right? The key to a successful

8:48

security strategy is layered security, but

8:50

it's also information. knowing

8:53

what's what's out there, what's going on, what

8:55

the threats are, what's happening in your network,

8:58

making decisions based on stale

9:00

data, trying to defend your

9:02

critical assets from cyber attacks with

9:04

tools that that don't even talk to each other.

9:06

That's don't wait for IT teams to

9:08

navigate today's attack surface. It's time for a

9:10

different approach. Tanium is one

9:12

of these disruptors that has come

9:14

along, burst into the and

9:16

has transformed everything. Tanium

9:18

says it's time for a convergence.

9:21

of tools, of endpoints,

9:23

and IT operations and

9:25

security. Now they have solutions

9:27

for every sec. government entities, education,

9:30

financial services, retail,

9:32

health care, you could trust

9:34

their solutions for every workflow that

9:36

relies on endpoint data, they've

9:38

got asset discovery and

9:40

inventory in Leo

9:43

fast, which means you could track down every

9:45

asset in your entire IT

9:47

space and and and know what you

9:49

own instantaneously. They'll help

9:51

you with risk and compliance management. They'll

9:53

let you find and fix vulnerabilities in

9:56

seconds at scale. Notice

9:58

there's a there's a little theme here at

10:00

scale. very fast. Right? Their

10:02

threat hunting is amazing. Hunt

10:04

for sophisticated adversaries in

10:06

real time. You

10:08

can do client management, automate

10:10

operations from discovery to management,

10:12

again, across your entire estate. You've

10:14

got sensitive data monitoring,

10:16

which is Laporte, you gotta index and

10:18

monitor your sensitive data. You could do

10:20

it globally seconds, you know, where every

10:22

bit of data is

10:24

and who has access to it, even maybe more

10:26

importantly. Tanie and protects organizations

10:29

where other endpoint management and

10:31

security providers have failed. all in

10:33

one platform. Tinuum identifies where

10:36

your data is across

10:38

all your entire IT

10:40

estate Can patch every

10:42

device you own in seconds. Can

10:44

implement critical security controls, and it could that do

10:46

it all from a single pane of

10:48

glass. Just ask Kevin Bush He's vice

10:50

president of IT at Ring Power Corp.

10:52

He says, quote, Tanian brings

10:54

visibility to one screen for our

10:56

whole team. And if you don't have that

10:58

kind of visibility, not gonna be able to sleep

11:00

at night. Sounds like Kevin

11:02

knows what he's talking about. With

11:04

real time data comes real

11:06

time impact If

11:08

you are ready to unite operations

11:10

and security teams with a single source

11:12

of truth and confidently protect

11:14

your organization from cyber threats, it's

11:16

time you met tanium. To

11:18

learn more visit tanium, TANIUM

11:21

tanium dot com slash

11:23

twit. Tinuum. dot

11:25

com slash tweet. We thank you so

11:27

much. For supporting SecurityNow,

11:29

you support us by using that address. They

11:31

know you saw it here at tedium dot

11:33

com. slash trent.

11:35

I'm ready my friend for the

11:37

picture of the week. So

11:40

for those who are not video

11:42

connected here. As always, I have

11:44

to explain this. We

11:46

have a a

11:48

flatbed transport

11:51

vehicle, a flatboard trailer sort of

11:53

thing. And it looks like there there's

11:55

a on the left, is

11:57

Laporte of a rust colored red

12:00

container where someone

12:03

probably said, hey, we we need you

12:05

to pick up some dirt So,

12:07

you know, bring a container

12:09

and we got some dirt for you.

12:11

Yes. Well, apparently,

12:13

the container that they brought

12:15

was too small. Yeah. Cons because it's,

12:17

you know, it's about one third of the

12:19

back of this of this

12:21

trailer, this flatbed. And

12:23

the rest of it has been piled

12:25

up with the overflow dirt

12:27

that didn't fit in the container. Now

12:31

in a saying world,

12:33

they would

12:34

throw a tarp over this whole

12:36

thing. Right? And like, lock the tarp

12:38

down, but Maybe

12:41

they didn't have a tarp. Anyway,

12:43

some apparent

12:45

rocket scientist here decided well,

12:47

you know, I've gotta do something. Right? because

12:49

I just got this exposed dirt

12:52

on the big pile on the back

12:54

of this trailer.

12:57

So they did

12:59

what they used what they had. They threw

13:01

a about a two inch diameter

13:03

belt. across the top

13:05

of the pile of dirt, which

13:07

is about I

13:09

don't know. It it

13:11

covers maybe four

13:13

percent of the of

13:15

the pile -- Mhmm. -- the rest of it

13:17

exposed to the air. Now there's

13:19

you can sort of see off on one

13:22

side Leo facing us. There is like there's looks

13:24

like there the strap was somewhere else initially

13:26

because you can sort see some of dirt was

13:28

flattened on the side there. Yeah. It's supposed to be a little

13:30

bit. So it looks like well, it looks like maybe the

13:32

strap was originally anchored on the

13:34

on the slot in the trailer one

13:37

notch further forward. CS Fed

13:39

points down. Too much time looking. So

13:41

there was yes. There was

13:44

Are are you

13:47

telling me there's a reason SpinRite took three

13:49

years? No. I think

13:51

Logan five in our chat room may have

13:53

come up with something. It's not to prevent

13:55

slippage. It's to prevent theft.

13:59

Oh, it's brilliant. You wouldn't want

14:01

someone to steal your dirt. don't don't

14:03

feel a certain man. So the so

14:05

this is like that poll that we

14:07

saw that had the bike lock around.

14:09

Right. where it didn't it TWiT it indicated

14:12

an intention without

14:14

actually providing any enforcement. I

14:16

love it. Anyway,

14:18

Oh, my. Oh, my. Once again,

14:20

we're we're we seem to be drifting here a

14:22

little bit off of the security related

14:24

topics related. The security dirt. No. No.

14:26

It's clear. Well, part

14:28

of the part of the goal of

14:30

podcast is to have some fun, and

14:32

so we're we're providing some entertainment to

14:34

TWiT. Yes. Okay.

14:36

With Firefox version 107

14:38

which was released last Tuesday a

14:40

week ago, nothing was

14:42

earth shattering. There was no

14:45

critical security fixes, but

14:47

there were a very large and

14:49

welcome collection of high severity

14:51

things fixed. No zero

14:54

days that were noted. There were

14:56

also a couple moderate severity

14:58

repairs. So you know,

15:00

it appeared to be primarily be

15:02

released just to fix those things

15:04

since there were not otherwise or even a

15:06

large number of new features a couple little

15:08

developer things. You know, they're

15:10

continuing to push the standards

15:12

which Firefox supports forward

15:14

because, you know, the web people can't keep

15:16

their hands off of like, oh, how

15:18

about if we added the ability

15:20

for it to read your mind?

15:22

That would be good. It's like okay, we don't

15:24

have that technology. You know, but let's develop

15:26

an API for that so that when we

15:28

do, web pages will be I mean,

15:30

that's this is what's going on. So a

15:32

little bit more of that is happening,

15:35

nothing else to see. the

15:38

It was interesting to me

15:40

to see that Google recently

15:43

settled something that we discussed four

15:45

years ago. This

15:48

was a suit brought

15:50

against Google by forty

15:52

states' attorneys general.

15:56

They settled for three hundred and ninety

15:58

one point five million dollars.

16:00

Where that number came

16:02

from? Oh, yeah. Only the

16:04

attorneys know.

16:06

As I said, we talked about this four years

16:08

ago back in twenty eighteen when

16:10

these offices of those forty Steve

16:13

general sued Google alleging

16:16

that Google had been lying and

16:18

misleading users into thinking

16:20

that they had disabled location

16:22

tracking in their account

16:24

settings. The

16:27

lawsuit followed some reporting that

16:29

was produced by the associated

16:31

press which found that Google was

16:33

continuing to track its users

16:36

even after they had enabled the

16:38

account privacy setting that

16:40

claimed to turn off location

16:42

tracking. So in that

16:44

settlement, Google agreed to pay

16:46

this three hundred ninety one and a half

16:48

million dollars in rest tune and

16:50

also, of course, to change the

16:52

way it handled location tracking

16:54

in the future. The the

16:56

first thing we're reminded of is

16:58

that the wheels of justice, when

17:00

they don't completely fall off the wagon,

17:02

do tend to turn slowly, at

17:04

least in the United States So

17:06

it took us four years to get to this

17:09

point. The other thing

17:11

we learned is that thanks to

17:13

Google's posting about this, their

17:15

own posting, we

17:16

learned what

17:17

has changed since then. So

17:19

their posting last

17:22

week was titled managing

17:24

your location data. And

17:26

it brings new meaning to

17:28

the phrase putting on a

17:30

happy face that

17:33

that they wrote Location information

17:36

lets us offer

17:38

you a more helpful experience

17:41

when you use our products From

17:43

Google Maps, driving directions

17:45

that show you how to avoid

17:47

traffic to Google Search,

17:49

surfacing local restaurants, and letting you

17:51

know how busy they are, you know,

17:54

like all the benefits, right, of Google

17:56

knowing where you are. They said

17:58

location information

18:00

helps connect experiences across

18:02

Google to what's most relevant

18:04

and useful. And,

18:06

okay, yeah, that's certainly the

18:08

case. or can be. They said

18:11

over the past few years, right,

18:13

while while this lawsuit was in the

18:15

works, we've introduced more

18:18

transparency and tools

18:20

to help you manage your data and

18:22

minimize the data we collect. That's

18:24

why we, and then they have three things.

18:27

launched Auto Delete Controls,

18:29

a first in the industry,

18:31

and turned them on by default

18:33

for all new users

18:36

giving you the ability to automatically

18:38

delete data on a rolling

18:40

basis Leo only

18:42

keep three months, eighteen

18:44

months, or thirty six months, worth

18:46

of data at a time. And if that sounds familiar

18:48

to our listeners just because, yes, we covered this

18:50

when this was happening. Second thing

18:53

they did developed easy to

18:56

understand things. I'm sorry.

18:58

Easy to understand settings.

19:00

like incognito mode on Google

19:02

Maps, preventing searches or

19:04

places you navigate to,

19:07

from being saved to your account.

19:10

And third, introduce more

19:12

transparency tools, including

19:14

your data in maps and search,

19:16

which lets you quickly access your key location

19:19

settings right from our core

19:21

products. And they said,

19:23

these are just some ways that

19:25

we to provide more choice and

19:28

transparency. Consistent

19:31

with these improvements, we

19:33

settled at investigation

19:36

TWiT forty US state

19:38

attorneys general based

19:41

on outdated outdated

19:44

product policies that

19:46

we changed years ago,

19:48

as well as okay.

19:51

In addition to the three hundred ninety one and a

19:53

half million dollars, outdated product

19:56

policies we changed years ago, as

19:58

well as a financial

20:00

settlement, We will be making

20:02

updates in the coming months to

20:04

provide even greater controls and

20:06

transparency over

20:08

location data. So things to

20:10

come. These updates include

20:12

three things: revamping

20:15

user information hubs To

20:18

help maintain how location data

20:20

improves our services, we're adding

20:22

additional disclosures to our

20:24

activity controls and

20:26

data and privacy pages. We're

20:28

also creating a single comprehensive

20:31

information hub that highlights

20:33

key location settings to help people make

20:35

informed choices about their data. Okay?

20:38

So, you know, more

20:40

transparency. Second thing

20:42

simplified deletion of location

20:45

data. We'll provide a new

20:47

control that allows users to

20:49

easily turn off their location history

20:52

and web and app activity settings

20:54

and delete their past data

20:56

in one simple flow.

20:59

will also continue deleting location history

21:01

data for users who have not recently

21:04

contributed contributed

21:06

new location data history to their

21:09

account. And third, updated

21:11

the account setup. will give

21:14

users setting up new

21:16

accounts. A more detailed

21:18

explanation of what

21:20

web and app activity is.

21:23

what information it includes, and

21:25

how it helps their Google experience.

21:28

So they finished today's settlement

21:30

is another step along

21:32

the path of giving more meaningful

21:34

choices and minimizing data

21:37

collection providing more helpful

21:39

services. So It

21:42

seems clear that what

21:44

was going on during these four

21:46

years, I mean, you know, lots of

21:48

back and forth was some

21:50

negotiation about the

21:52

things that that

21:54

Google was being asked to

21:56

do proactively in

21:58

order to make what, you

21:59

know, make what they were

22:01

doing, make this tracking behavior, which

22:04

initially got them into such trouble,

22:07

that, you know, these forties attorneys general decided to

22:09

gang up and and say,

22:11

look, this needs to change.

22:14

So you know, stepping back from this

22:16

a bit, it it must be

22:18

truer than than I

22:20

guess I'm able to understand that

22:23

The more information an advertiser has

22:25

about someone, the more revenue

22:27

is generated by showing

22:30

that person advertisements you

22:32

know, I mean, as our list as our listeners

22:35

know, I've always been somewhat skeptical

22:37

about that. I mean, that it that it can

22:40

mean that much, but Yeah.

22:42

It seems to me that advertisers would not be

22:44

trying so hard if it

22:46

didn't really make them more money.

22:49

since they also know that no one

22:51

wants to be profiled and tracked across

22:53

the Internet. So they wouldn't be

22:55

risking our wrath to the degree they

22:57

are if it really really wasn't

23:00

valuable to them. So

23:02

anyway, we've got

23:04

well, in a minute, we'll be talking about

23:07

a different issue with some

23:09

more attorneys general and the

23:11

FTC.

23:14

I caught wind of

23:16

a mention that Red

23:18

Hat had started cryptographically

23:21

signing its deployment zip

23:23

files. Oh, god. Can I have made

23:25

me curious since I'd

23:27

never heard of zip files being

23:29

cryptographically signed. We've always

23:31

talking about executables being

23:33

signed. And, you

23:36

know, we know that web web

23:40

assertions of their identity are

23:42

signed, but That was new for me for

23:44

zips. And with all

23:46

the problems that we've been seeing with

23:48

supply chain poisoning, obtaining

23:51

verifiable assurance of

23:53

an archived unmodified

23:56

authenticity, that would be

23:58

great. So A

23:59

cryptographic

23:59

signature could do that.

24:02

And cryptographic signing

24:05

wakes

24:05

weeks

24:06

makes way more sense

24:09

than the old school practice

24:11

of publishing the hashes

24:14

of files on the same

24:16

site where the files are

24:18

being hosted for download. Doing

24:21

that never made any sense

24:23

to me. since if a bad guy was

24:25

able to compromise a web

24:27

server to alter the

24:29

files being downloaded from

24:31

that site, What

24:33

would keep them from also updating the

24:35

hashes shown at the

24:37

same site as proof of a

24:39

file's authenticity? and

24:41

I'll talk about a false sense of

24:43

security. So, anyway,

24:45

this is a lot better than that.

24:47

So looked into what was going on, and I

24:50

found a posting by Red Hat

24:52

titled cryptographic signatures

24:54

for Zip Distributions. I

24:57

paraphrased what they posted to remove a lot of

24:59

their oversimplified descriptions for our

25:02

audience. So they

25:04

wrote Our build

25:06

system, Brew, produces

25:09

our RPM and ZIP

25:11

distributions, and automatically hash

25:13

the archives it makes.

25:16

The hashes are

25:18

used to validate that the files

25:20

have not changed before

25:22

they're uploaded to our CDN

25:24

and made available to customers.

25:26

We've taken advantage of this aspect

25:28

of our build process,

25:30

and extended it by

25:33

combining all of the

25:35

hashes for a particular release,

25:38

and packaging them into

25:40

an SHA2 fifty six

25:43

sum file. So SHA2

25:45

fifty six SUV. file. This

25:48

file is in a standard format

25:50

that lists the hash and

25:52

the corresponding the corresponding

25:55

file name of the

25:57

particular file artifact

25:59

as is a term they use.

26:01

It is commonly used across the

26:03

industry to provide integrity to binary

26:06

files. However, it's not

26:08

limited to that. The

26:10

SHA two fifty six sum

26:12

command on Red Hat enterprise

26:14

Linux, other Linux distributions,

26:16

and Mac OS natively

26:19

support this file format. They

26:22

said since our software production

26:24

team has completed their

26:26

verification procedures I'm

26:28

sorry. Once our software production

26:30

team has completed their verification

26:33

procedures. They sign off on the

26:35

release from both a process and

26:37

technical perspective. the SHA two

26:39

fifty six some file they

26:41

created is signed by

26:43

our latest release Leo,

26:45

which produces a dot

26:48

ASC file.

26:50

This file is an

26:52

Asci armor formatted

26:55

detached signature file

26:57

that proves the integrity and

27:00

provenance of the SHA two fifty

27:02

six thumb file and

27:05

Transitively, the zip

27:07

file artifacts enumerated

27:09

within that file. The

27:12

gPG command on Red Hat

27:14

Enterprise Linux, other Linux

27:16

distributions and macOS supports the

27:18

file format natively. due

27:21

to the potential damage that a

27:23

lost or stolen private key

27:27

could cause we've taken additional

27:29

steps to add assurance to

27:31

the signatures we produce.

27:33

The primary technology behind this

27:35

is our signing server.

27:37

To sign these files, we

27:40

use a high strength ninety

27:43

six bit private key. And

27:45

our public keys are available

27:47

on our website and

27:49

the MIT, you know, Massachusetts

27:52

Institute of Technology Public

27:54

key server. Okay. So

27:56

that's what they posted.

27:59

Red Hat's mention of

28:02

a detached Signature simply

28:05

means that the signature itself

28:07

resides in a separate file.

28:09

The signature is just a

28:12

an an sha two

28:14

fifty six hash of the

28:16

file it's signing which is

28:18

then encrypted under Red

28:21

Hat's Super Secret, and in

28:23

this case, very long, forty

28:25

ninety six bit private key

28:28

which they're careful not to let

28:30

loose. You know, just

28:32

like my GRC code signing

28:35

keys, it probably resides in a NHSM,

28:37

a hardware security module

28:40

where it literally cannot be extracted.

28:42

It can only be used.

28:44

So there's no reason

28:46

for that signature file not

28:49

to stand alone. That is again,

28:51

it's just you so

28:53

there's this this composite file,

28:57

which contains this

29:00

the the hashes and

29:02

the files that they that

29:04

they were hashed from,

29:07

that's just AAA

29:09

list a textual listing,

29:11

an ascii file.

29:13

That file is then SHA2

29:16

fifty six hashed. That's

29:19

the file you whose integrity you wanna

29:21

verify. That SHJ two

29:23

fifty six hash is then

29:25

signed with their

29:27

with their signing

29:29

server and and and

29:32

there was and which

29:34

is to say that the the SHJ two

29:36

fifty six hash is encrypted

29:39

with the private key.

29:42

So that creates an

29:44

encrypted blob, which is the

29:46

signature, and it's a

29:48

freestanding file. So

29:50

somebody who then wants to verify

29:52

that, uses the Red

29:55

Hat's private key, which is

29:57

available from several sources, so you don't have

29:59

to worry about that being

30:01

screwed TWiT, in order

30:03

to decrypt the blob, that

30:05

will bring that decrypting that

30:07

blob which restores the

30:09

SHA two fifty six hash, which you

30:11

can then use to verify

30:13

that the file of

30:15

the hashes that you've got

30:18

matches and has not been tampered with.

30:22

So, you know, this is

30:24

a welcome move as

30:26

a deterrent to the abuses that we

30:28

are now seeing and talking about more

30:30

and more of today's supply

30:33

chain And it's probably

30:35

where the broader open

30:37

source community will need to go.

30:40

The glitch here the glitch to

30:42

doing that is that Red

30:44

Hat Enterprise the enterprise linux

30:48

Corporation, you know, Red Hat Corporation

30:50

has no problem maintaining a

30:52

siding signing server and

30:55

buying a certificate that asserts their

30:57

identity. But the open

30:59

source world has always had a

31:01

problem with the need to pay

31:03

for certificates. as we know,

31:05

let's encrypt, solve this

31:07

problem by making TLS certificates

31:09

free for web servers.

31:12

But the challenge here is

31:14

not the same. Let's

31:16

encrypt offers no

31:18

guarantees about the

31:20

identity of a TWiT

31:23

provides domain validation

31:26

certificates where the only requirement is

31:28

for the certificate to match the

31:30

server's domain name. Specifically,

31:33

it does not offer that is

31:35

Let's encrypt, does not offer

31:37

O. V. Organization

31:40

validation certificates. in order to

31:42

issue o v certificates, any

31:44

certificate authority must

31:47

by by universal agreement

31:50

perform some significant

31:52

reconnaissance to positively

31:54

verify the identity of the entity

31:56

requesting the certificate so that

31:58

the of Venus means

32:00

something. And what's more? Of

32:03

course, many open source projects

32:05

are just some guy working alone

32:07

without any organization to

32:09

be validated. So maybe

32:11

the solution will be, for

32:14

example, to come up with

32:16

a secure means for

32:18

submitting repositories to GitHub

32:21

for its signing

32:23

with its signature then

32:27

using some much stronger

32:29

means for asserting

32:31

the identity of the individual

32:34

requesting the signing service. For

32:36

example, that process might

32:38

require much more

32:40

rigorous multi factor authentication, something

32:43

again, you're you're really wanting

32:45

to put it out of the reach of bad

32:47

guys to get in there and screw this up

32:49

so that it means something. So it's a problem that

32:51

needs to be solved. But, you know, one

32:53

way or another, we need a solution to

32:55

this current supply chain

32:57

pollution problem And, you

32:59

know, the application of a bit of

33:01

of a bit of crypto might be a

33:03

place to start. So, you know, hats off

33:05

to Red Hat for doing a

33:07

little pioneering here in in that way.

33:10

Okay. Now,

33:15

the FBI purchased

33:18

Pegasus. You know, that's the NSO

33:20

group's infamous smartphone

33:25

spyware platform. They said it

33:27

was for, quote, research and development

33:30

purposes, unquote. Last

33:33

week, Yeah. What are they developing? I wonder.

33:36

Uh-huh. Yeah. Last week, the

33:38

New York Times ran a story with headline

33:41

internal documents show

33:43

how close the FBI came

33:46

to deploying spyware. Now

33:50

I have a little bit different take on this, but we'll get to that

33:52

a second. The New York

33:55

Times reported that

33:57

last December. FBI

33:59

director Christopher Ray told

34:02

Congress this is behind this is a

34:04

closed door testimony that the bureau purchased

34:07

bureau as a new Federal Bureau

34:08

of Investigation, the bureau purchased

34:12

the infamous Pigasus phone hacking tool for

34:15

research and development purposes.

34:18

Well, it turns out

34:20

that Voya the US

34:22

Freedom of Information Act can

34:24

be quite handy for figuring

34:26

out things that really

34:28

happened. Here's how the

34:30

times explained what they found. They wrote,

34:32

during a closed door session with

34:34

lawmakers last December, Christopher

34:38

a Ray spelled WRAY

34:41

the director of the FBI, was

34:43

asked whether the bureau had

34:46

ever purchased and

34:48

used Pegasus. These are

34:50

like directly asked. The hacking

34:52

tool writes the times that

34:55

penetrates mobile phones and

34:57

extracts their contents. Mister Rae acknowledged that

35:00

the FBI had bought

35:02

a license for

35:04

Pegasus, but only for

35:06

research and development, quote,

35:08

to be able to figure out

35:10

how bad guys could use

35:12

TWiT. For example, He

35:16

told senator Ron Wyden,

35:18

according to a transcript of the hearing

35:20

that was recently declassified. But dozens

35:23

of internal FBI documents and court records

35:25

tell a different story

35:27

writes The Times, The

35:29

documents produced in response to a Freedom of

35:32

Information Act lawsuit brought by The New York

35:34

Times against the bureau

35:36

show that

35:38

FBI officials made a push in late twenty twenty and

35:40

the first half of twenty twenty

35:42

one to deploy the

35:44

hacking tools made by the

35:46

Israeli spyware firm NSA

35:48

in its own criminal investigations

35:50

that is in the FBI's

35:53

own criminal investigations The officials developed advanced

35:56

plans to brief the bureau's

35:58

leadership and drew up

35:59

guidelines for

36:02

federal prosecutors about how the

36:04

FBI's use of hacking

36:06

tools would need to be

36:08

disclosed during criminal

36:10

proceedings, like Okay? How

36:12

how did you get this information?

36:15

Well, it came

36:18

to us. Uh-huh. So the Times writes it's unclear

36:20

how the bureau was contemplating

36:22

using Pegasus and whether

36:24

it was considering hack the

36:27

phones of American citizens,

36:30

foreigners, or both. In

36:32

January, the Times revealed that

36:34

FBI officials had also

36:36

tested the NSO tool

36:38

Phantom, a version of Pecos'

36:40

capable of hacking phones with

36:42

US numbers. The FBI

36:44

eventually decided not to

36:46

deploy Pecos in criminal

36:48

investigations in July of

36:50

twenty twenty one. amid a

36:52

flurry of stories about

36:54

how about how the hacking tool

36:56

had been abused by

36:58

governments across

37:00

the globe. But the documents offer a glimpse

37:02

at how the US government

37:04

over two

37:06

presidential administrations wrestled

37:08

with the promise and peril

37:10

of a powerful cyber

37:12

weapon. And despite the

37:14

FBI decision not to use Pegasus,

37:17

Court documents indicate the bureau remains

37:20

interested in potentially

37:22

using spyware in future investigations.

37:26

Okay. And of course,

37:28

the Times reporting brings up

37:30

the question of Christopher Rae's

37:34

apparently misleading testimony

37:36

in front of congress,

37:38

senator Ron Wyden is

37:40

not is not happy about that.

37:42

In a statement from his it read, it

37:45

is totally unacceptable for the

37:48

FBI director to

37:50

provide miss leading

37:52

testimony about the bureau's

37:54

acquisition of powerful hacking

37:56

tools and then wait months

37:58

to give the full story

37:59

to congress and the American

38:03

unquote. So

38:05

the times revealed in

38:07

January that the FBI

38:09

had purchased Pegas' in twenty eighteen and over

38:11

the next two years tested the

38:14

spyware at a secret

38:16

facility in New Jersey. Since

38:20

the bureau first purchased the

38:22

tool, it has paid approximately

38:24

five million dollars

38:26

to the NSO group. Now,

38:29

it seems to me that the issue

38:31

with Pecosys is less

38:33

about its use

38:36

then it's potential for misuse and abuse.

38:39

The worry is that

38:41

once they have

38:44

it, Repressive governments would be unable to resist

38:46

the temptation of using it to spy

38:48

on political rivals. We'll see an

38:50

example of that here in a moment.

38:54

and of course, dissidents and other non criminal

38:58

actors. And of course, Pecos'

39:00

doesn't respect

39:02

geopolitical boundaries. So

39:04

anyone who has it can aim it at anyone

39:07

else anywhere. But in

39:09

the United States, we have

39:12

a system For obtaining court orders, for

39:14

searching, and for making

39:16

legal within bounds,

39:18

what would otherwise be illegal

39:22

reconnaissance. So as long

39:24

as the FBI would

39:26

only be using Pegas' within

39:29

our constitutional protections, I think that

39:31

it would be a useful tool

39:33

to empower their criminal

39:36

investigations and yes,

39:38

they would be required to tell a judge that

39:40

this is what we want to do,

39:42

this is how we're gonna do

39:45

it, and we have probable cause and all

39:48

the other, you know,

39:50

requirements of getting a

39:52

court order to

39:54

pursue things like a wiretap and so forth.

39:56

So TWiT seems

39:58

to me, the Yes.

40:00

It is problematical

40:02

because it could be abused.

40:06

But if we're gonna have systems

40:08

that are

40:10

that that are otherwise not prone

40:14

to be subject to court

40:16

order search

40:18

then maybe this is the way it happens. Yeah. I mean

40:20

-- Okay. -- I mean, we allow wire taps

40:22

under quarter three. Exactly.

40:26

Is this Is Peguis

40:28

somehow too dangerous to be

40:30

used? I think the

40:32

concern is TWiT just

40:34

its control. that all the

40:36

reports we have suggest that it

40:38

is a zero click tool --

40:40

Right. --

40:42

which TWiT it is possible

40:44

to target at an individual smartphone

40:48

and it

40:50

goes in against all of the attempts by Apple

40:52

and Google, you know,

40:54

iOS and Android keep

40:56

it out. there

40:58

are enough ways in that it gets TWiT, and

41:00

then it's able to provide the

41:03

entity that deployed it

41:05

with information, you

41:07

know, the the equivalent of someone

41:10

unlocking their phone and

41:12

also being eavesdropped on,

41:14

it's able you know, it is a

41:16

surveillance tool. I guess, question always is

41:18

is it is it first of all, it's gonna be

41:20

a very expensive. It's a million dollar

41:22

surveillance tool. Right? It's very

41:24

very expensive. multi. Multi.

41:26

Because it can't be used too often or it

41:28

loses its its usefulness.

41:30

Because as soon as

41:32

the company's you know, find it. They'll they'll defend against

41:34

it. So these these zero days are

41:36

very, very expensive, especially And it

41:38

might click. also might very

41:40

well be that it is tightly

41:42

tethered. Well, it that's actually

41:44

and this might have been the problem. As I

41:46

understand it, The NSS0

41:49

group is responsible for the

41:51

hack. You don't just give the

41:53

FBI pegasus to say, have fun with

41:55

it, guys. Yes. It doesn't work that way.

41:57

Right? Right. Right. So that's another problem

41:59

is that some international company,

42:01

an Israeli company, would

42:03

then be privy to what you're

42:06

doing. Right? Yeah. That might be a bigger

42:08

problem. And other entities may

42:10

not care but that may be

42:12

something that we can't, you know, get over. And

42:14

in fact, maybe that was the

42:16

beginning, you know, the you know,

42:18

in in

42:20

testimony like this. There's typically some piece of truth.

42:22

So probably yeah.

42:24

The FBI said, maybe

42:26

we need to be empowered

42:28

with this

42:30

tool because we're unable to get in any other way.

42:32

So let's buy a copy

42:34

and let's learn how it

42:37

works. Let's have the Pegas' experience

42:39

Gibson that we can decide if this is something

42:42

that, you know, we can sell to the

42:44

greater government. It's

42:46

my understanding that what you're buying

42:48

really is the NSO

42:50

group access. They

42:52

trigger it on let's say they wanna

42:54

on my phone. The NS0 group gets into my phone triggers it and

42:56

then hands control over to the

42:59

FBI. Right. So

43:02

Right. I'm sure that's

43:04

illegal in the US because

43:06

that's that's an Israeli in

43:09

a company, not even a government entity,

43:11

but a but a business, that

43:13

your the FBI says, okay. Well, wanna

43:15

hack Leo's phone. Here's his phone. Hack it

43:17

for us. I don't I that

43:19

can't be legal. some NBA. Right? They just it

43:21

literally was research. They just wanted, well, Leo know, let's let's understand it

43:23

a little bit. But I can't imagine the NSO

43:26

giving the keys to the Kingdom of the

43:28

FBI either. That's why they do

43:30

it that way. Right? No.

43:32

Yes. And and in fact, I have a an

43:34

another related story that sort of

43:36

speaks to

43:38

that. Greece that, you

43:40

know, that Athens government,

43:42

Greece, bought a a related

43:44

program predator for

43:46

seven million euros. Wow.

43:49

A

43:50

recent report in the Greek press

43:52

claimed that Greece's government

43:54

paid seven million euros to

43:58

Intelexa, INTELLEXA

44:00

Intelexa,

44:02

for access to the predator surveillance

44:04

and spyware platform, and an

44:07

additional hundred and fifty thousand

44:09

euros for the ability

44:12

to rotate ten new

44:14

targets per

44:16

month. So that says, yes,

44:18

they were not given, you know,

44:20

carte blanche. They had to, you

44:22

know, their it is

44:24

tightly tethered to under

44:28

Intelexus control.

44:30

So This little

44:32

bit of accounting news follows

44:34

the massive scandal of the

44:36

Greek government having been caught using

44:40

the spyware to go

44:42

after, not only rival

44:44

political parties, but also

44:46

journalists and prosecutors investigating

44:50

government corruption. So this is

44:51

the this is the double edged sword,

44:52

is that, you know, it it

44:54

seems to be impossible

44:56

for governments that purchase this

45:00

to to behave themselves. Well, again,

45:03

I would hope that if it

45:05

were made possible for the FBI

45:07

to acquire this technology TWiT

45:10

would be done above board. It would be done within the

45:13

constitutional protections of the government. I'm sure

45:15

there were those, Hutlino,

45:18

Edward Snowden, who don't

45:20

believe could be

45:22

possible. But we do have we've set up a

45:24

situation where

45:26

where The technology that our private citizens corporations

45:28

are using is not subject

45:30

to court orders. And

45:35

you know, thus the tension that

45:38

we're currently under. So,

45:40

anyway, it's again, as I said, it

45:42

seems to be the problem is less about the

45:44

tool -- Mhmm. -- than how it's used.

45:46

Mhmm. You know, it is

45:48

technology. It already exists

45:50

and it's going to

45:52

exist. So makes more

45:54

sense to me to properly

45:56

regulate and control its use than

45:58

to attempt to deny

46:00

it completely. which, you know, just forces its use underground. And

46:02

maybe it's old fashioned, but I also feel

46:04

like, where the

46:05

United States, we should

46:08

be better than those other guys. You know, I

46:10

agree. You know, we should have higher

46:12

standards. Just just because

46:14

other countries use these tools doesn't

46:16

mean we have to. Yeah.

46:19

I agree. Leo, I also

46:21

wanna breathe. Okay. I

46:24

do. You agree that it's time to

46:26

take a

46:28

break. You and I both agree about

46:30

that. Our show today brought to you by Barracuda.

46:32

We love Barracuda. Barracuda.

46:36

is a security company that we use, we work with, and a

46:38

lot of people should be working WiPeep. They

46:40

have done some really interesting research.

46:44

They have a threat, you know, a really, really high

46:46

quality threat team. And

46:48

they Steve identified thirteen

46:50

kinds of email threats that are

46:52

in widespread use.

46:54

Cybercriminals are using them every day. Some of them

46:57

you know, fishing, spear phishing, conversation

47:00

hacking, ransomware,

47:04

total of ten plus

47:06

three. So thirteen tricks,

47:08

the three I've just mentioned and and ten

47:10

more. that cyber criminals use

47:12

to steal money from you, from your company,

47:14

or personal information from

47:17

your employees and customers. So now the

47:19

question every business owner should be asking is, we safe? Are

47:21

we protected? I ask Brussels

47:23

every time I

47:26

see him. email cybercrime. It is

47:28

probably the number one way

47:30

bad guys get into

47:32

your system. and it's

47:34

becoming more sophisticated, attacks

47:36

are getting harder and harder to prevent

47:38

perimeter defenses are often

47:40

insufficient. They're gonna use social

47:42

engineering, you know, fear and

47:44

and urgency to convince

47:46

your employees to do something

47:48

they want and to do. Social

47:50

engineering attacks, including spear phishing and Business

47:53

email comprise, cost businesses on average a hundred

47:55

thirty thousand dollars an instant.

47:57

A hundred thirty thousand dollars

47:59

an instant as demands for

48:02

COVID nineteen tests, for instance, at the beginning of a twenty

48:04

twenty two rose. Of course, what

48:07

happens? Barracuda's researchers

48:10

saw an increase in COVID nineteen test related phishing

48:12

attacks. Between October and

48:15

January of this year, five

48:17

hundred twenty one percent increase. Because the bad

48:19

guys, you know, they're they're watching the

48:22

headlines. Again, fear and urgency.

48:24

Right? They're gonna pray on your

48:26

weaknesses. When everybody got

48:28

really interested in cryptocurrency in the

48:30

late twenty twenty, and I guess

48:32

they're interested now for other reasons

48:34

now, but but a year ago,

48:36

remember, and there were all these ads

48:38

encouraged. If fortune favors the brave and

48:40

all that, the price of Bitcoin went up four

48:42

hundred percent between October twenty twenty,

48:44

April twenty twenty one, Guess what?

48:46

Barracuda Research found that

48:49

impersonation attacks using

48:52

Bitcoin and crypto as the,

48:54

you know, The come on increased a hundred and ninety two

48:56

percent in the same period. The

48:58

Internet crime complaint center last

49:00

year, the

49:02

IC three seem nineteen thousand three hundred sixty nine,

49:04

business email compromise, and

49:06

an email account

49:08

compromise complaints. I'm

49:10

sure that's just the tip of the iceberg. That's the ones those are

49:12

the ones that came into the IC three with adjusted losses of

49:15

over one point eight billion

49:18

dollars.

49:19

It's not enough to

49:20

secure your email at the gateway. The perimeter

49:22

defense not gonna do it. You

49:24

got of course, you gotta have you

49:26

gotta have gateway security to protect against,

49:29

you know, malware viruses, zero days,

49:31

all that stuff, spam. You

49:33

gotta fight that too. But your gateway

49:35

is defenseless against spear phishing against

49:37

targeted attacks, attacks. you know,

49:39

that seemed to come from the

49:42

boss, you know, to an

49:44

employee by name,

49:46

for instance. Protection at the inbox

49:48

level. And and by the way, this has to

49:50

include AI and machine learning.

49:52

Is necessary

49:54

necessary to detect and stop the most sophisticated

49:56

threats. I

49:57

get a solution for you.

49:58

It's very easy. Just get the copy of the

50:00

Barracuda report. You should be reading Laporte. Your

50:03

IT department should be reading this. Thirteen email

50:06

threat types to know about right

50:08

now. It explains how cyber criminals are

50:10

getting more and more sophisticated every day

50:12

how you could build the best protection for your

50:14

business, your data, and your people, and of course,

50:16

do it with Barracuda. Find out

50:18

about those thirteen email threat

50:20

types you need to know about, and

50:22

how Barracuda can provide

50:24

complete email protection for your teams,

50:26

your customers, and your reputation. Get

50:28

your free ebook at

50:30

barracuda dot com slash security

50:33

now barracuda dot

50:35

com slash security now.

50:38

Barracuda, your journey. secured.

50:40

We thank him so much for supporting

50:42

Twitter and especially for supporting

50:45

now and Steve's work. They

50:47

care a lot. about your security and they know Steve is here to make

50:49

a big difference. You help us, by the way, when you

50:52

use that address always with all of these

50:54

ads, go to barracuda

50:56

dot com, slash

50:58

security. Now we thank you,

51:00

Barrick. Otherwise, they don't know why they're suddenly

51:02

getting so much more busy than they are.

51:04

I mean, we needed to know.

51:07

It's us. It's us, maybe. Steve,

51:10

Steve. Here we go. The

51:14

password manager one password has added support

51:16

for pass keys to

51:18

its offering,

51:20

and In a nice promotion

51:22

of pass keys, they've created

51:24

a community supported online

51:27

directory listing online

51:30

services currently supporting authentication.

51:32

I've been waiting for this because I

51:34

want to play with passkeys. You know,

51:38

I've got iOS one point one point or sixteen point one

51:40

point one, I think, now. And

51:42

it's supposed to support passkeys, but

51:44

I've never tried it. So now

51:46

we can. So this directory

51:48

is at pass keys

51:50

dot directory. I didn't know

51:52

there was a directory was a TLD,

51:54

really. They've just patent out of

51:56

control Leo. Is there a

51:58

dot Leo? Leo probably is. Anyway

51:59

so again, passkey's

52:02

dot directory takes you to

52:05

this listing. It currently

52:07

has forty three companies listed with

52:09

their URLs, although some

52:12

are flagged

52:14

as MFA. So, you know, multi factor authentication. So I

52:16

suspect that they might not be

52:18

pure passkey's login. They

52:20

may be passkey's plus another

52:22

factor, which

52:23

would be annoying. So anyway, some notable

52:26

names on the list, which do

52:28

appear to be pure passkey's

52:30

authentication without that

52:32

MFA tag. include

52:34

a one password passkey's

52:37

demo page of all

52:39

Leo Buy. It's

52:42

best time. yeah, supports basket

52:44

carnival cruises -- Good. -- eBay -- Good.

52:48

-- kayak. you

52:50

know, the the travel site,

52:52

microsoft dot com. Again, nest cafe,

52:55

like what? Sure. Why

52:58

not? NVIDIA,

53:02

PayPal, and Robinhood. So

53:04

anyway, I just discovered this as I

53:07

was putting the the the the podcast

53:09

together. So I have not made any time to experiment with and

53:11

explore, but I am an avid

53:13

buyer on eBay.

53:16

Oh. often buying,

53:18

like, old hard drives that I need

53:20

to make sure that spin right works

53:22

with. Or in fact, I'll be talking

53:24

about spin right in a in a few minutes here because I actually did

53:26

just buy four drives from eBay,

53:29

which were specific drives

53:31

that I needed. So anyway, I

53:34

ought to be able to give logging in

53:36

to eBay. I've passed keys. I think

53:38

I I'm seeing it. Let me let me

53:40

log in and I'll show you. I'm

53:42

gonna log in to I'll go to Carnival Cruises, and it says

53:44

create create an account. And

53:49

let me show you this. I'm gonna make bigger and

53:51

see that log in

53:53

with your phones face

53:55

ID or fingerprint. That's pass keys.

53:58

It may not say pass keys.

53:59

Right? Right. Right. Right.

54:02

Right. So

54:04

scan this QR code. Alright. Let me try it with your phone's

54:06

camera. So this is yeah.

54:08

That's cool. I'm so glad this is

54:10

the first time I've seen it.

54:14

Alright. I scan it with my camera. I'm logging in.

54:16

Enter your email. Okay.

54:18

I always do it. And the site

54:20

knows, Leah. Look what it's doing. Oh, it

54:23

does. it knew I did something. Yeah. How would

54:26

it know that?

54:28

Well, because You're

54:31

I'm going to a special URL that that that

54:33

that QR code. Okay. Connection

54:36

loss. Something went wrong. Try again.

54:38

I'll, crud. Well, well, after all, it is

54:40

carnival cruise. So We're

54:42

working working

54:44

on it. So

54:47

now what do I do? Now what do I do to take another picture? Let's do it again. Did

54:49

you all I can't. Do you do you

54:51

have an account at Carnival? Not

54:55

carnival. No. Okay. Not I

54:57

mean, how about kayak? Well,

55:00

I think the idea is you would

55:02

have to should I go somewhere? I

55:04

already have an account? You wanna see what that

55:06

That looks like. I don't know. Let

55:08

me let me just I didn't

55:10

do it quickly enough probably. me What happens if

55:12

I log out of eBay? because

55:15

I'm, like, statically. Yeah.

55:17

So now I'm

55:20

pressing continue. Do you oh,

55:22

here it is. Do you wanna allow carnival dot

55:24

com to use face ID?

55:26

Continue. I'm using face

55:28

ID. The the it worked.

55:30

And look at this, on the phone, it

55:33

now says Steve if I could find

55:35

that. Password lists sign

55:38

in enabled. enable evil Fast

55:40

log in by own ID, but this

55:43

is pass keys. Right?

55:46

You've

55:46

got to be

55:47

pass keys. Yeah. So

55:50

that's cool. So now it's

55:52

once, you know, complete your profile blah blah

55:54

blah. But now

55:56

I presume From now on, I can just use my phone. love

55:58

it. Yay.

55:59

Yes.

56:01

Yes. Now

56:03

I have carnival cruise line log in.

56:06

Yeah. Sure. I want

56:08

that. No. Actually, the cruise line that

56:10

we do go on is owned by them. So

56:13

I guess that's one of the cruise lines we'd like to go

56:15

on. Again, I this is the weirdest

56:18

list, like Best Buy -- Well,

56:20

Carnival Cruises. I don't know. But

56:22

don't you think it's like I should trust these people

56:24

because they're at least on top of it.

56:26

Chase Chase is not there, b of a

56:28

is not there. Well, banks is gonna

56:30

be a higher standard. You know, if you in Nescafe, but not Starbucks.

56:32

It's like, okay. I don't know

56:34

what's going on. But anyway,

56:37

I think it's gonna be lower stakes companies. Don't you

56:40

think initially a bank? That's gonna be

56:42

problematic, problematic. Yeah. I guess

56:44

Microsoft has become lower

56:46

stakes here. Would I like to receive

56:48

emails? No. Do I

56:50

accept their terms and

56:52

conditions? Yes? Have

56:54

you already booked your cruise? No. Okay. Now

56:57

now I guess the next time

56:59

I go there, let's go on another

57:01

let's go on another computer.

57:04

This is this is Steve first time I've ever used this. That's cool. So now I'm

57:06

gonna say log in and

57:08

it's gonna say log in with

57:10

your phones face

57:12

ID or fingerprint. I'm gonna

57:14

click that. Oh, I

57:16

have to scan it again. Is that is that right? Is that

57:18

what it should be doing? Yes. Yes. Because don't you

57:20

you haven't transferred your passkey into that computer.

57:22

And then it says, do you wanna log in

57:24

using a saved account? Yes. Logging

57:28

in Bingo, highly o, bunch of

57:30

Leo. It

57:33

works. It's a

57:35

little onerous. It's So

57:38

will I always have to scan my QR code to get in? So

57:41

well, so what you're doing is you're using

57:43

your phone's pass key -- Right.

57:45

-- in order in order to

57:47

authenticate across to a different device. Right. You and

57:50

and this was the problem that I

57:52

talked about is that

57:54

is that

57:56

if, you know, squirrel would there would only be one.

57:58

But but so you need to create another

58:00

pass key in your laptop.

58:03

And so there is there should

58:06

be a way to to you you can't export

58:08

the passkey, but you can you

58:12

can link them, you

58:14

you can create another passkey and then link

58:16

them so that they're identified as the

58:18

same. So yeah. See, I don't see I

58:20

already have a Microsoft account, but I don't see any

58:22

way to log

58:24

in if you haven't set up PASKIs

58:26

with PASKI. Right? I'm just

58:28

going to the Microsoft site. Now I do

58:30

have an account and I could sign in, but I'm gonna say, could I do this with

58:33

my passkey? No. But maybe if I go

58:35

into my account, I could what's

58:37

that little thing down at the bottom? Sign in options,

58:39

but that's I already looked at that, and it just

58:41

gives me GitHub or forgot my

58:44

username. That's not That's

58:46

not passkey. I bet you I have to go to the

58:48

Microsoft account. They'll log

58:50

in normally and then say and I would

58:52

like to establish passkeys

58:54

with this probably. Yeah. Yeah. That would make sense. Yeah. I'll

58:56

try it while you're talking. Anyway,

58:58

all of our listeners now, again,

59:00

pass keys

59:02

dot directory you can check back there and maybe

59:04

eventually some more interesting

59:06

sites will be available. Right. I think it's TWiT

59:08

miracle that Robinhood is using it.

59:12

Yeah. Yeah. Yeah. I bet FTX would

59:14

have if they Go go go to pass keys

59:16

dot directory and see what it Oh, oh,

59:18

there's more

59:20

than just this. Okay. Yeah. Yeah. Yeah. Just yeah. Yeah. because there was a bunch of

59:22

things that also had MFA for

59:24

tags for some reason. So

59:26

so you can see the the little the little

59:28

green dots just in.

59:30

Sign in. And here's Cloudflare MFA.

59:33

Yeah. So I didn't know what

59:35

that meant. probably means I need a password

59:37

to log in and pass

59:39

keys. like it's two factored. Yeah. That's what

59:41

I'm thinking. Just DocuSigning. I could sign

59:43

in Oregon. Interesting. Well, I

59:46

have a GitHub account.

59:48

Let me Leo me play with

59:50

that a little bit and see. Oh, okay. Cool. Yeah. In other

59:53

news, again, paseke's

59:55

dot directory, our listeners, Okay.

59:58

So from the having fun with bureaucracy

1:00:02

department comes in edict

1:00:04

from the

1:00:06

OMB. The US's office

1:00:08

of management and budget has

1:00:10

ordered federal agencies to scan

1:00:12

their systems. Oh, yes. Scan

1:00:15

Those puppies carefully scan

1:00:17

them in scan them

1:00:19

and provide an inventory

1:00:21

of assets containing cryptographic

1:00:24

systems that could be

1:00:26

cracked by quantum computers in

1:00:28

the coming years. That's it.

1:00:31

How would you know? Well,

1:00:33

Leo. Okay. First of all, there is probably not a

1:00:35

single computer in the government

1:00:37

that doesn't use and

1:00:40

depend upon some public

1:00:42

key crypto, and none

1:00:44

of the currently deployed public

1:00:46

key crypto. There's no West Quarter

1:00:49

is quadrant resistant. Yeah. So the

1:00:51

0MB could have

1:00:53

simply said, give us a list of

1:00:55

all your computers. That's a good point. And and by the way,

1:00:57

stop using them. Yeah. That okay.

1:01:00

So the the next point worth

1:01:02

noting is just a

1:01:04

reminder that No

1:01:06

one has come near

1:01:08

to building a quantum computer

1:01:10

anywhere, so far as anyone knows,

1:01:13

they could even begin to

1:01:16

think about breaking

1:01:18

actual public key crypto.

1:01:21

Oh, yes. Factoring the number twenty seven, we

1:01:23

can do that. It's magic.

1:01:26

But the number thirty

1:01:28

five, we're not quite

1:01:30

there yet. Gibson us another ten years or so, and we'll be able to

1:01:32

factor thirty five. Okay. Now that

1:01:34

said, I'm on

1:01:36

the record agreeing

1:01:38

that there's absolutely no reason

1:01:41

not to move us

1:01:43

to Quantum Safe Crypto sooner

1:01:46

rather than later, you know. Let's not wait till

1:01:48

we need it because we know how slow

1:01:50

and painful these moves can

1:01:52

be. So you know, just

1:01:54

as sure or just as

1:01:56

soon as we're absolutely

1:01:58

sure that we're not

1:01:59

gonna be making a

1:02:02

big mistake. because that's possible. Remember that one

1:02:04

of the candidates that had

1:02:06

already been chosen,

1:02:08

already selected, was

1:02:10

recently cracked by conventional computers.

1:02:13

So it would be a

1:02:15

lot better that,

1:02:17

you know, without for us to

1:02:20

stay where we are, where we know we can't crack today,

1:02:22

the the the algorithms

1:02:26

we're using, before moving prematurely to

1:02:28

something that we presume

1:02:30

some future non

1:02:32

existent mythical

1:02:34

quantum computer should also be unable

1:02:37

to crack. So

1:02:39

the OMB eDICT

1:02:40

stated

1:02:42

that federal agencies had until May fourth

1:02:45

twenty twenty three. So, like, is

1:02:47

that, you know, this coming

1:02:50

May fourth. I don't know why

1:02:52

May fourth, but that's

1:02:54

it. And the NSA

1:02:56

ordered that all government

1:02:58

agencies handling

1:03:00

classified information must use quantum

1:03:01

resistant encryption by thousand

1:03:04

and thirty five. Okay?

1:03:06

So that's thirteen years from now.

1:03:10

By then, we ought to be up to

1:03:12

factoring forty five. So

1:03:17

Good to be we'll be switching to over

1:03:19

to quantum computers any minute, you know,

1:03:21

before we need

1:03:24

them. Yeah. Okay.

1:03:26

So this other piece

1:03:28

of attorney's general news that

1:03:30

I wanted to share, one

1:03:33

of the developing themes of

1:03:35

this podcast is the observation that we're still in

1:03:37

the Wild West stage of

1:03:39

the creation of

1:03:42

the Internet. it

1:03:44

remains an unregulated

1:03:46

or only very loosely regulated

1:03:48

medium. And of course,

1:03:51

globally, it's a uncoordinated total disaster.

1:03:54

The idea that we've

1:03:56

linked our fundamentally our

1:03:59

fundamental insecure networks to

1:04:02

those of openly hostile nations

1:04:05

should give anyone pause.

1:04:08

Yet, that's what

1:04:10

we've done. Chinese, Russian, and

1:04:12

Iranian cyber criminals under the

1:04:14

protection of their nation

1:04:16

states who have no love for

1:04:18

the US. are able to

1:04:20

openly attack the networks

1:04:22

of US corporations and its

1:04:24

private citizens. And yes,

1:04:26

there's reciprocity is

1:04:28

able to do the same to them and presumably

1:04:30

that's happening too, although there seems

1:04:32

to be a surprising lack of information

1:04:34

about that. You know, But,

1:04:37

you know, reciprocity doesn't make any of this

1:04:40

sane. You know? It's like,

1:04:42

you know, mutually assured

1:04:44

destruction. So we can

1:04:46

only hope that the

1:04:48

Internet our grandchildren will

1:04:50

use as adults, thirty

1:04:52

years from now, will be much different

1:04:54

from the one we've been watching being born

1:04:56

through these past thirty years.

1:04:59

I bring this up because

1:05:01

various democracies around the world, notably the

1:05:04

EU and the US,

1:05:06

among others, are inching

1:05:08

forward cautiously in an attempt to

1:05:10

provide their citizens with some

1:05:12

legally enforceable rights

1:05:14

to privacy and personal information.

1:05:17

At the moment, We have clear statutes

1:05:19

outlawing overt network intrusion and

1:05:22

attack. When those laws are

1:05:24

broken, people lose their freedom for

1:05:26

doing so. but

1:05:28

nothing yet prevents or

1:05:30

regulates the passive collection

1:05:32

of as much Internet user

1:05:35

data as possible. Google was sued by those forty

1:05:37

states' attorneys general, not for

1:05:40

tracking, but for tracking

1:05:42

after they said they weren't.

1:05:45

As long as a company

1:05:47

doesn't say that they won't do something,

1:05:49

they can do pretty much anything they

1:05:51

want. So how do we get

1:05:53

this to change? Here's a hopeful example.

1:05:55

Last Thursday, a coalition of thirty

1:05:58

three state attorneys general

1:05:59

co signed a letter

1:06:02

formally urging

1:06:04

the US Federal Trade Commission, RFTC,

1:06:06

to pass legislation which

1:06:09

would regulate online data

1:06:12

collection practices. might

1:06:14

not happen, but it's a good

1:06:16

start. These AGs said

1:06:18

they are, quote, concerned

1:06:20

about the alarming amount of sensitive

1:06:22

consumer data that is amassed, manipulated, and

1:06:25

monetized, unquote. And

1:06:28

they also said that

1:06:30

they regularly Steve inquiries

1:06:32

from consumers within their states

1:06:34

about how their own data

1:06:36

is being hoarded and abused. Okay.

1:06:39

So since we've still got a bit of time, and I think

1:06:41

this is extremely important, I'm gonna

1:06:43

first share just

1:06:46

the introduction in

1:06:48

the letter, which was submitted to the FTC

1:06:51

and signed. It's really

1:06:53

pretty. They have, like, different

1:06:55

colors of ink on the signatures. I don't know how they pulled

1:06:57

this off, but it was, like, you know, signed by

1:07:00

forty Steve attorneys

1:07:02

general. So in their in

1:07:04

the beginning of this letter, they said,

1:07:06

we, the attorneys general

1:07:09

of Massachusetts, can I'm not gonna read

1:07:11

them all because they didn't list them all,

1:07:13

but they they did some. Massachusetts, Connecticut,

1:07:16

Illinois, New Jersey, North Carolina, and

1:07:18

Oregon joined by

1:07:20

their respect of attorneys general of the undersigned

1:07:22

states right to the Federal Trade

1:07:24

Commission in response to the August

1:07:26

twenty two

1:07:29

twenty twenty two advanced

1:07:32

notice of proposed rulemaking

1:07:35

on commercial surveillance

1:07:37

and data security. So this

1:07:40

was something that the FTC put

1:07:42

out there and asked

1:07:44

for comments. So that was propose

1:07:47

an advanced notice of proposed rulemaking on

1:07:49

commercial surveillance and data security.

1:07:51

That all sounds

1:07:54

great. So they said, as the chief consumer

1:07:56

protection officials in most of our

1:07:58

respective states, we hope

1:08:00

to inform

1:08:02

the commission as it

1:08:04

contemplates new trade regulation

1:08:06

rules governing commercial surveillance

1:08:09

and data security. The

1:08:11

state attorney's general commend the

1:08:13

FTC for its comprehensive review of corporate

1:08:16

surveillance and data security

1:08:18

in preparing the notice.

1:08:20

We, too,

1:08:23

are concerned about the alarming amount of sensitive consumer data

1:08:25

that is amassed, manipulated,

1:08:27

and monetized. Our

1:08:30

offices frequently receive outreach

1:08:32

from consumers concerned about

1:08:35

the privacy and security of their information. Research supports that

1:08:39

consumers are worried about commercial surveillance and feel

1:08:42

powerless to address it. Oh, really?

1:08:45

Leo. Imagine

1:08:48

that. That's interesting. We're just going on the record

1:08:50

here. Many consumers believe that tracking by companies

1:08:56

is inevitable. yet often do not

1:08:58

even know what is being recorded. These fears intensify

1:09:01

when they learn

1:09:04

more about the commercial surveillance

1:09:06

economy, and in particular consumers fear falling victim to

1:09:09

identity theft and

1:09:12

data misuse. A majority doubt

1:09:14

that their data can be kept secure. Contributing to these

1:09:16

concerns is the

1:09:19

fact that companies

1:09:21

that companies are often collecting more data

1:09:23

than they can effectively manage or need

1:09:26

to perform their services. Our

1:09:31

consumer privacy related enforcement actions

1:09:34

and investigations have resulted

1:09:36

in settlements that work

1:09:39

like Google that have provided significant business practice changes

1:09:41

to strengthen data security

1:09:43

and privacy going

1:09:46

forward. but there is still more work to be done. Our

1:09:49

submission highlights the heightened

1:09:52

sensitivity of certain

1:09:54

categories of consumer information. the

1:09:57

dilemma of data brokers and how they prevail

1:09:59

consumers and

1:09:59

how

1:09:59

data minimization can

1:10:03

help mitigate concerns surrounding

1:10:07

data aggregation. Okay. Then

1:10:09

the letter goes on at

1:10:11

quite some length detailing

1:10:13

five general categories of abuse. Unfortunately, in an effort to be

1:10:15

very clear and to drive their points home, that

1:10:17

part is too long

1:10:20

to share. but

1:10:23

I found a separate release about this

1:10:25

action from New Mexico's

1:10:28

attorney general,

1:10:31

Hector Baldares. In it addressed each of these

1:10:34

five points by reference quite succinctly. So those I wanna share

1:10:36

because It's

1:10:39

good stuff. So first, there there

1:10:41

so there's five five categories.

1:10:44

Location location data data.

1:10:46

He said, or his office said, according to the letter,

1:10:49

many consumers are not even aware of

1:10:51

their location that their

1:10:54

location information is being

1:10:56

collected. and when a consumer

1:10:58

wishes to disable location sharing, their options are quite limited. The attorneys general

1:11:03

sensitive of this information, which can

1:11:06

reveal intimate details of daily

1:11:08

life, such as

1:11:10

where they live and their

1:11:12

shopping habits, their daily schedule, or where they

1:11:14

visited the doc or whether they visited the doctor

1:11:18

or pharmacy. laws passed in states like California, Connecticut,

1:11:21

Virginia respect the use

1:11:23

and collection of

1:11:26

location data can provide a framework to inform the

1:11:28

FTC through the rulemaking process.

1:11:31

So this is this is

1:11:33

him saying or his office

1:11:35

saying, look at what For

1:11:37

location data things, look at what California, Connecticut, and

1:11:39

Virginia have done, use that, you

1:11:41

know, consider using that

1:11:43

as a framework. biometric

1:11:48

data. The coalition urges the

1:11:50

FTC to consider the risks

1:11:52

of commercial surveillance practices

1:11:55

that use or facilitate the use

1:11:57

of facial recognition, fingerprinting, or other

1:11:59

biometric technologies. Many consumers provide this

1:12:01

information to companies for

1:12:04

security purposes or

1:12:06

to learn about their ancestry. But consumers are not always made aware when

1:12:09

their data

1:12:12

is collected how it

1:12:14

is used or if it is resold for purposes to which never meaningfully

1:12:16

consented.

1:12:20

Medical data. The FTC should also

1:12:22

consider the risks of practices that

1:12:27

use medical data regardless of whether the

1:12:30

data is subject to the health insurance portability and accountability

1:12:33

act of nineteen

1:12:36

ninety six. popularly known

1:12:38

as HIPAA, and the privacy rule. Medical data not

1:12:40

necessarily covered by HIPAA

1:12:43

is referred to as health

1:12:47

adjacent data, which can

1:12:49

be collected by many devices.

1:12:51

For instance, smartwatches, health monitors,

1:12:53

sleep monitors, and health or wellness phone

1:12:56

applications. The letter also

1:12:58

highlights medical information risks

1:13:00

through examples such as storage

1:13:02

of health related Internet searches

1:13:04

or appointment scheduling information

1:13:07

being passed to others

1:13:09

through online tracker tools. In other words, you get a sense

1:13:11

for how comprehensive this letter was that the forty states attorney's

1:13:14

general submitted to the

1:13:16

FTC. the middle Two

1:13:19

more to go, data brokers. The attorneys

1:13:21

general reiterated to the

1:13:24

FTC the persistent

1:13:26

dangers of data brokers. data

1:13:29

brokers provide consumers I'm sorry. Data brought

1:13:31

brokers profile consumers by

1:13:36

scouring social media profiles, Internet

1:13:38

browsing history, purchase history, credit

1:13:42

card information, and government

1:13:44

records like driver's licenses, census

1:13:47

data, birth certificates, marriage licenses, and voter

1:13:52

registration information. Data brokers use

1:13:54

this information to create profiles of certain consumers,

1:13:56

which can be

1:13:59

purchased by almost anyone. based

1:14:02

on susceptibility to certain advertising

1:14:05

or likelihood to buy

1:14:07

certain products. This

1:14:09

scale of aggregation of anonymously gathered

1:14:12

information Leo identify consumers

1:14:14

and put consumers at

1:14:16

risk of

1:14:18

scams unwanted and persistent advertising, identity

1:14:21

theft, and lack of

1:14:23

consumer trust in

1:14:26

the websites they visit. And

1:14:28

lastly, data minimization. The attorneys general say that

1:14:30

it is vital that the FTC consider data

1:14:35

minimization requirements, and limitations. With respect

1:14:38

to data collection and retention,

1:14:41

the letter encourages the FTC to examine the

1:14:44

approach taken in

1:14:46

California, Colorado, Connecticut,

1:14:50

Utah, and Virginia consumer privacy which

1:14:52

mandate that businesses tie

1:14:54

and limit the collection

1:14:56

of personal data

1:14:59

to what is reasonably necessary. In relation

1:15:02

to specified purposes, limiting the collection

1:15:04

and retention of

1:15:07

data by businesses will improve

1:15:09

consumer data security as businesses will have less data to protect and

1:15:11

less data potentially available

1:15:15

to bad actors.

1:15:17

Okay. So I think if nothing else, this is

1:15:19

a useful start. In the

1:15:22

United States where we exalt

1:15:24

capitalism, one

1:15:27

to innovation, but we all know that we're

1:15:30

a long way from being, you know,

1:15:33

from being endanger

1:15:36

of that. Much of what is going

1:15:38

on today is only able to happen under the cover of darkness because

1:15:41

consumers are

1:15:44

blissfully unaware You know,

1:15:46

what did Apple discover when they started requiring their apps to proactively obtain cross

1:15:48

application tracking

1:15:52

permission? They found that

1:15:54

nearly everyone who was asked declined. No thanks. You know,

1:15:58

and no surprise. So

1:16:01

We can expect any improvements to be slow going. As I always say, change is

1:16:03

slow, but the pressure

1:16:08

is there. and it's not

1:16:10

gonna go away. At least I think we're moving in the right direction. And, you know, this,

1:16:12

you know, forty

1:16:15

states getting behind this you

1:16:18

know, one wonders why it's not

1:16:21

fifty.

1:16:21

well Well,

1:16:22

knows who knows?

1:16:24

Some presumably buckled

1:16:26

to some pressure. Okay. I have some closing the loop things that I think are

1:16:31

interesting. Vincent shot

1:16:34

me a note that I wanted to share

1:16:36

regarding we we were

1:16:38

talking about the concern that was

1:16:41

raised by a different listener about

1:16:43

the Xima board and how when he changed his credentials, it was

1:16:45

only away from the

1:16:48

log Leo of

1:16:51

Casa OS. Casa OS. It was only

1:16:53

for the web portal log on,

1:16:55

and all of the

1:16:57

other credentials remain the same. He

1:16:59

was concerned that that that would

1:17:01

that the lack of

1:17:04

changing

1:17:04

of other

1:17:07

credentials was unknown to Xima board users and that

1:17:10

they might get themselves in trouble, for example, if they turn this thing into a router.

1:17:12

Anyways, Vincent Stacey

1:17:15

tweeted, Steve, PF

1:17:18

Sense installs its own version of Linux and won't have the default

1:17:20

users of

1:17:24

another distribution.

1:17:25

And that's a very good point for

1:17:28

router. Though just

1:17:30

for the record, it's

1:17:32

actually free

1:17:34

BSD units that PFS runs on top of and brings along with it.

1:17:36

But the main reason

1:17:39

why a Xima board would

1:17:43

not be my first choice as a router, is that

1:17:46

unless a network expansion board

1:17:48

were to be plugged

1:17:50

into its PCIe by four

1:17:52

slot, TWiT only

1:17:54

has a pair of land nicks built in. And I would expect a

1:17:59

router today certainly

1:18:01

one that any of our

1:18:03

listeners would be using to have a few more interface controllers,

1:18:06

a few more nicks.

1:18:09

for implementing useful multi network isolation. So I

1:18:11

can't see it being a

1:18:16

big being really popular as

1:18:18

a router. There are some, you know, some better,

1:18:20

you know, fan

1:18:22

less

1:18:23

solutions like that what

1:18:26

is it? The SG1 thousand, I

1:18:28

think, that that

1:18:30

that I've talked about before.

1:18:33

Charles Turner tweeted

1:18:35

as possible fodder for a listener feedback

1:18:37

section in a future episode

1:18:39

of SecurityNow podcast.

1:18:41

I have a question arising from the

1:18:43

discussion you and Leo had on Tuesday okay.

1:18:45

He says November fifteenth. That was last

1:18:48

Tuesday. During

1:18:50

security now, episode eight, safe languages.

1:18:52

Yep. Last last podcast. He says,

1:18:54

with the future of Twitter in

1:18:58

doubt, What is your prediction on the long

1:19:01

range fate of Mastodon? The

1:19:03

cynical part of me gives

1:19:05

Twitter a fifty fifty chance

1:19:07

of either a rebounding back

1:19:10

to its former glory as or and beyond or

1:19:12

b

1:19:13

becoming a forty

1:19:15

four billion version billion

1:19:19

dollar version next iteration

1:19:22

of MySpace and FTX.

1:19:24

okay so Okay.

1:19:26

So It's clear to us all

1:19:28

that Twitter is currently in

1:19:30

turmoil, and I don't have

1:19:32

any firsthand sense. for just

1:19:34

how fragile Twitter's technology is internally. And it

1:19:37

seems to me that

1:19:39

matters a lot. If

1:19:42

the previous regime engineered really solid bulletproof systems, then

1:19:45

in order to be

1:19:47

able to withstand Elon's shaking

1:19:51

of its foundation. But overall, I'm

1:19:53

a big believer in

1:19:56

inertia and in

1:19:58

things generally changing much more slowly than we expect. Leo, of

1:20:01

course, Elon could trip over

1:20:03

the main power chord and

1:20:06

Twitter could go dark until someone plugged it

1:20:09

back in. And I suppose I'm

1:20:11

interested in what Elon is

1:20:13

doing there. You know? He's an interesting character,

1:20:16

and somehow he he's managed to

1:20:18

get other people in the past, at

1:20:20

least, to do

1:20:23

some truly amazing things. I'll never forget

1:20:25

the sight of those twin booster rides returning

1:20:27

to and landing on that floating platform

1:20:32

for reuse. That was truly astonishing

1:20:34

-- Yeah. -- technology. And it's Elon's SpaceX

1:20:37

StarLink technology, which

1:20:40

actually works that's enabling

1:20:42

Ukraine to survive Russia's increasingly aggressive attacks against its

1:20:48

infrastructure. Again,

1:20:50

Thanks, Elon. Mostly

1:20:52

though, my take is that I

1:20:54

think Elon is just having

1:20:56

fun with his life. as

1:20:58

is his as is his

1:21:01

right. Right? You know? I hope

1:21:03

he's had expensive fun. You

1:21:06

know? What about our lives? He

1:21:08

doesn't care. He doesn't care.

1:21:11

No. He doesn't. He thinks

1:21:13

resimulations. That's why. It's it's his

1:21:15

life. Yeah. And he's not a guy

1:21:17

who likes to make small waves.

1:21:20

Right. Elon's

1:21:22

waves are big. and let's not forget

1:21:24

that Twitter made him

1:21:27

do it. They insisted that

1:21:30

he honor his wildly overpriced purchase offer.

1:21:33

He didn't wanna buy Twitter.

1:21:35

They made him buy it. So

1:21:37

it seems to me that Twitter is

1:21:39

getting what it deserves. The Elon

1:21:42

treatment. He's showing them

1:21:44

that he can do anything

1:21:46

he wants to with it. So all

1:21:49

of this made me curious about what he

1:21:51

is doing with it. You know,

1:21:53

I pick up little bits

1:21:55

here and there but I don't follow

1:21:57

news feeds or even Twitter because they interrupt my work and my train

1:21:59

of thought. So it

1:22:02

was with some joy. that

1:22:05

I stumbled upon a site which I figured

1:22:07

had to exist somewhere. The site's called, Twitter

1:22:10

is going great dot

1:22:12

com. in

1:22:15

the spirit of Molly White. Yep. And,

1:22:17

yes, of course, it's

1:22:19

offering up its

1:22:21

share of Shouten Freud So keep in mind that

1:22:24

it's naturally gonna be biased, but

1:22:26

it's still a lot of fun.

1:22:28

The site hosts

1:22:31

a simple timeline of Twitter's Elon

1:22:33

related happenings. So now I can check-in from

1:22:36

time to time whatever

1:22:38

I want to, you know,

1:22:40

to get a sense for what's going

1:22:43

on over there. I mentioned it because I imagined that some of our listeners would also appreciate knowing

1:22:46

about this nicely distilled

1:22:50

timeline, event resource. It's

1:22:52

hysterical because it's all tweets. Yes. I guess

1:22:54

that's that's a best source of what's

1:22:56

going on at Twitter, I guess.

1:22:59

Yeah. Twitter is going great dot com. I'll

1:23:01

show you another one that you

1:23:03

should read. This is

1:23:06

from a Twitter Reliability site

1:23:08

reliability engineer. Mhmm. I think former

1:23:11

Matthew Tayo, he's on substack, and

1:23:16

it's I think you would enjoy

1:23:18

this. I barely understood it, but he talks about all of the redundancies, all

1:23:20

of the automation. He says,

1:23:23

when I came in, The

1:23:26

list of servers was on a spreadsheet.

1:23:28

Now, of course, it's a much better

1:23:31

system, and he did a

1:23:33

really good job. It sounds like he and

1:23:35

his team did a really good job of

1:23:37

making it run. He was in charge

1:23:39

of the cash, the

1:23:41

cash team, which was a pretty

1:23:43

big deal because everything you're getting is

1:23:45

served from cash. None of it served

1:23:47

from Leo served. I I

1:23:49

don't didn't wanna interrupt you, but

1:23:51

Does Eddie has Eddie would stop to think

1:23:54

about what it does?

1:23:57

Oh, it's phenomenal. Yeah. It is un

1:23:59

Well, freaking believable. This

1:24:01

is just a fraction of

1:24:03

it. What Twitter actually Leo.

1:24:06

Yeah. I I can't imagine building

1:24:08

this system. Oh, yeah. It just stopped. Read this, so

1:24:10

I think you'd enjoy it. And it's just a fraction of what

1:24:12

is going on. And

1:24:15

but his point is These

1:24:18

things are designed to run unattended. A

1:24:20

lot we automated everything

1:24:22

we could. And so it should

1:24:25

unless something you know, nobody's

1:24:28

gonna kick the plug out of the

1:24:30

side. I hope there's more than one

1:24:32

plug. But Good afternoon. You piss off. You piss off. You

1:24:34

let me you might Well, that might yeah. He might he just might pull the blood. So but

1:24:39

but you wouldn't expect it to all fail all of a sudden.

1:24:41

There may be bugs here and there and stuff. And there may and the real problem is there may not be somebody to solve

1:24:43

that problem, which cascades

1:24:47

another one. etcetera. Then I've read I've read a number of

1:24:49

articles. We have Phil Livingston, who was the

1:24:51

founder of Evernote. Very, very

1:24:53

well rounded. I

1:24:55

was very impressed a Sunday. He's a smart guy. And

1:24:57

-- Yeah. -- he was saying, you know, give you on as you

1:25:00

do, just

1:25:02

give you on some credit. There was a good article by a former Tesla engineer

1:25:04

that says Elon did exactly the same

1:25:06

thing in twenty eighteen to Tesla.

1:25:10

He was firing people. His spent in the nights there. He was bemoaning. There

1:25:12

might be bankrupt. This was all in the

1:25:14

lead up to the type three.

1:25:17

The model three

1:25:19

of the Tesla. And so

1:25:21

this is kinda how Elon works, obviously, you know, for some people not the ideal

1:25:23

situation. That's why have left Twitter

1:25:26

voluntarily as well as

1:25:28

involuntarily. Leo

1:25:31

I've also read articles who let's

1:25:33

say, you know, this is how he

1:25:35

he's reinventing Twitter. is

1:25:38

get you have to get rid of almost everybody and then

1:25:41

build a team of people who

1:25:43

believe in your vision. He

1:25:45

hasn't really communicated that apparently, but who believe in

1:25:47

your views. He's just making it up. Yeah. That's making

1:25:49

it up as he goes a lot. Nobody you

1:25:52

know, I don't I'm confused.

1:25:54

I see stuff that looks crazy. He says, you're gonna have a committee

1:25:56

to approve who who comes and goes, and then

1:25:58

he just says, now, I'm gonna bring him

1:26:00

back. And, you know, it's

1:26:02

just it's it's it's seems

1:26:04

chaotic. There was one there

1:26:06

was one piece there that that that said he sat down and explained

1:26:09

to the core

1:26:12

team how how advertising should

1:26:14

be tweets. And he and they said, they are.

1:26:16

He said native. Yeah. It

1:26:18

should be native. It is. That's

1:26:23

exactly my problem with the average. Yeah. He

1:26:25

you know, so he's coming somewhat from

1:26:27

ignorance, but you're right.

1:26:30

He's also a pretty interesting He's

1:26:32

probably sleeping there. He's there. He says he is. hours

1:26:34

a day and and, you know, he'll figure this thing out.

1:26:36

He and he he's a weirdo.

1:26:38

And some of the things he's tweeted.

1:26:42

I'm not thrilled about some of the pictures and stuff. This is from one twenty AM at Twitter.

1:26:44

He this is when he,

1:26:46

you know, came TWiT become in

1:26:48

Saturday. or

1:26:51

Friday night, Saturday morning, to explain

1:26:53

how Twitter works. And

1:26:56

these are the the the

1:26:58

skeleton crew there is sitting with

1:27:00

them but this is his picture of

1:27:02

what they drew on the whiteboard. This is not a code review. This is explaining

1:27:04

in rudimentary fashion to

1:27:07

somebody doesn't know how this

1:27:10

stuff works, how it's working. I get you know what? We don't know

1:27:12

yet. He may this may

1:27:14

be two hundred two point o.

1:27:18

inventing, and maybe this is how he works. I would

1:27:21

never wanna work for him, but

1:27:23

people And we'll see

1:27:25

what There was an interesting

1:27:27

moment I was watching a press conference when

1:27:31

Biden was off

1:27:35

in the east And it it

1:27:37

was that that awkward that awkward press conference where he

1:27:40

meant to

1:27:43

say Cambodia he said Columbia three times. like, oh, Joe. Joe.

1:27:45

But but but but someone in

1:27:47

the press pool asked

1:27:51

him about Elon. And

1:27:52

and so

1:27:53

understand that our relationship, the government

1:27:55

has a relationship. Right?

1:27:59

With Elon, because he's now SpaceX, and we got

1:28:01

all these contracts. Right. So

1:28:03

Biden just locked up.

1:28:05

You know, he it's didn't know

1:28:07

what to say because like, oh, you know, III

1:28:10

can't I don't dare piss off

1:28:14

Elon or, you know, We're gonna be in real we're not gonna have any

1:28:16

think we're looking into it though. You

1:28:18

know, it's it's very complicated because

1:28:21

Elon has relationships with not just the US government.

1:28:23

but many other governments Tesla sells a lot of cars and

1:28:26

builds them in China. It's a complicated

1:28:28

system. And it's kind of a bowl

1:28:30

in a China shop, but I

1:28:33

think he's a care I just think

1:28:35

he's a care person. I think he's having fun with

1:28:37

his life. Yeah. And, you know Too bad, though, because

1:28:39

observer. Twitter is a valuable resource

1:28:44

resource. It's not a public resource. It's

1:28:46

not even publicly held company anymore. And

1:28:48

it's incredibly valuable. Leo it's a

1:28:50

shame if he crashes TWiT, you

1:28:52

know? Well, I mean, I I he

1:28:54

he's I just he's led a bunch of loons back on recently. Yeah.

1:28:56

And, you know,

1:28:59

but I don't I don't ever see tweets from Loons.

1:29:01

I have a very quiet experience with Twitter. I just

1:29:03

talk to our listeners -- Right. --

1:29:05

and and they talk to me. And

1:29:08

it's just a great little channel.

1:29:10

So, you know, I don't care who says, you know, that that vaccines or

1:29:13

garbage alone. Yeah.

1:29:16

Who cares? Okay, Leslie McFarlane said,

1:29:18

hi, Steve. Uh-oh, if Twitter implodes, are you

1:29:20

are you going

1:29:23

to Mastodon or somewhere else.

1:29:25

Your SecurityNow podcast is top notch security and quality.

1:29:27

Well, thank you, Leslie. though

1:29:31

So okay. In order to

1:29:33

get the word out to eighteen years' worth of spin right owners,

1:29:36

I will shortly and I mentioned

1:29:38

this before on the podcast be setting

1:29:40

up an

1:29:42

old school email facility. One of

1:29:44

the several lists that I'll be maintaining

1:29:47

will be for security now listeners

1:29:49

who would like to subscribe to the

1:29:51

weekly links and the show notes and a description

1:29:53

of each week's podcast which I

1:29:55

post to Twitter. TWiT it'll

1:29:58

be nice to have more than two

1:29:59

hundred and eighty characters for that.

1:30:02

So so that will be a possibility.

1:30:04

And You know, as

1:30:06

for Mastodon, I have no I I'm took me

1:30:08

ten

1:30:09

years to get

1:30:11

Steve on Twitter.

1:30:14

Right. Patient. Thank you, Leo.

1:30:16

I'm I'm looking at it. I'm

1:30:18

not looking I'm not looking for

1:30:20

more connectivity. We'll see how Twitter

1:30:23

goes. As it is, I spend most

1:30:25

of my time in GRC's quiet

1:30:27

news group getting actual getting

1:30:29

actual work done. Yeah. and now we have Gitlab

1:30:31

for managing spin right bugs and feature

1:30:34

requests. And I have GRC's web

1:30:36

forums, which will

1:30:38

soon be quite active since that's where spin right's tech support will

1:30:40

be hosted. And a lot of

1:30:42

new users are gonna be using

1:30:44

spin right six one

1:30:47

and and have questions. or maybe not

1:30:49

because it's pretty much the same as

1:30:51

it was. It just works a lot better. You know, so anyway, I I

1:30:53

just don't have any

1:30:55

additional bandwidth available. for

1:30:57

new conversation opportunities. You know, I doubt that Twitter can actually implode. It's as

1:31:00

you said, Leo, it's too big

1:31:02

and too important. You know, I doubt

1:31:06

that even Elon can or will kill You

1:31:09

know, and I have an alternative

1:31:11

means for communicating my and

1:31:13

GRC's events to anyone

1:31:16

who cares through good old email.

1:31:18

So and I will extend this offer after episode 999

1:31:21

You can always

1:31:24

use us to tell the world, I

1:31:26

would bet a lot of spin right users and owners listen to various

1:31:30

other things we do. And We have a lot of different including Twitter channels.

1:31:32

So And Leo, we still have two

1:31:34

years. Who knows? Two years from two

1:31:37

two years from now.

1:31:39

What'll be going on? Okay.

1:31:41

Someone said, I where does his name go? Oh, I didn't have his name here.

1:31:43

Shoot. I think

1:31:47

it was Walt. Anyways, it's

1:31:50

Steve. Did you see? There's a project Hale Mary in

1:31:52

IMDB. project to hail mary

1:31:54

in i am db He

1:31:56

said crossing my fingers. Anyway, indeed,

1:31:59

there is AAA project Hale

1:32:04

Mary movie is in the works. Well,

1:32:06

it is currently flagged as in development. If you had listened to our interview with Andy Weir some

1:32:08

years some months ago when

1:32:10

it came out, he'd already optioned

1:32:12

it. And

1:32:14

he told me and I wasn't

1:32:16

too thrilled. I don't know how well I hit

1:32:18

my discomfort that Ryan Gosling had signed on

1:32:21

Yes. I saw that too.

1:32:24

I saw that. And I went,

1:32:26

oh, okay. But, yeah, we're gonna, you

1:32:28

know, Andy

1:32:31

was gonna be on some months ago,

1:32:33

but he just had a baby. We'll

1:32:35

get him back on. And

1:32:37

by the way, Daniel Suarez

1:32:40

has a new book. The

1:32:42

sequel to his delta book

1:32:44

is coming out soon

1:32:46

in, I think, next January. and we'll

1:32:48

have a lot of fun reading those. Those are great. Yeah.

1:32:50

So we'll get him on too. So, yeah, we'll we'll keep an eye. I'll

1:32:53

have Andy long before

1:32:55

a movie gets made. get the

1:32:57

latest on that one. Okay. So speaking of

1:32:59

books we've loved, so many people have written to me that telling me

1:33:01

that they've that they're

1:33:03

loving the silver ships

1:33:06

series that I wanna share a

1:33:09

tweet I received two days ago.

1:33:10

From the first person I

1:33:13

know who has or we know

1:33:15

who was finished the entire twenty

1:33:17

four book series. I was

1:33:20

horrified as I started to read

1:33:22

the tweet that he might have written

1:33:24

something of a spoiler, but that

1:33:26

concern was misplaced. So here's the content of the DM that Bob

1:33:29

Grant sent.

1:33:32

He

1:33:32

wrote, Wow wow

1:33:33

wow wow superb ending to the

1:33:35

series. There was enough great writing

1:33:37

and new intrigue in the first

1:33:39

part of this final

1:33:43

book in the Silver Ship's series to

1:33:45

be a great book in and

1:33:47

of itself. However, the

1:33:49

wrapping up of all the

1:33:51

various storylines from previous twenty three books. And

1:33:54

he says perez twenty silver

1:33:56

ships and the related

1:33:58

four Pairian books. At the

1:34:00

end, was

1:34:02

superb. There were joyful and poignant endings to each of the major characters

1:34:07

from the books. I

1:34:09

have to say that this is the best series I've

1:34:11

ever not to

1:34:15

take away from Weber's

1:34:17

honor verse. Of course, he's talking about David David Webber's, you know,

1:34:20

honor Harrington

1:34:24

series that was one of the

1:34:26

early series that we talked about in this podcast, or Rick Brown's frontiers series,

1:34:29

he says both of

1:34:31

which I've enjoyed. but

1:34:34

these twenty four books have been a joy to read from beginning to end.

1:34:36

And then he said after a little

1:34:38

break to catch up on some other reading,

1:34:43

I plan to start the new Scott Yuha series called

1:34:45

Gategoasts, whose first book is

1:34:48

Axis Crossing. And

1:34:51

as I mentioned to you, Leo, there are six more in

1:34:53

that series after these twenty

1:34:55

four. So anyway,

1:34:58

obviously, Bob has been following along with my previous

1:35:00

reading discoveries. He knows of

1:35:02

and read David Weber's Honor

1:35:04

verse series and Rick Brown's work

1:35:06

in progress from Tears Saga series. And, you know, for

1:35:08

what it's worth, I'm in complete agreement with

1:35:10

him about this being the best series

1:35:14

I've ever read. I'm at the start now nineteen of those

1:35:17

twenty four, so I have

1:35:19

six to go. And

1:35:21

having already made

1:35:23

this large investment in this series.

1:35:26

I'm delighted to learn in advance that it ends wonderfully. So

1:35:29

anyway,

1:35:30

one

1:35:30

last piece. Simon,

1:35:32

he

1:35:34

said, hi, Steve, persistence paid

1:35:37

off. I was able

1:35:39

to disable one time

1:35:41

code feature, he

1:35:43

has in quotes. He's talking about

1:35:45

PayPal. He said, you can call PayPal and

1:35:48

ask to

1:35:52

unconfirm your

1:35:52

phone number. It may impact

1:35:54

use of the PayPal app, but as long

1:35:56

as you do

1:35:59

not confirm

1:35:59

phone number, TWiT will

1:36:02

not text security codes. So that's Wait a minute. Which

1:36:04

is less secure.

1:36:06

Having no two factor or

1:36:08

having SMS

1:36:11

two factor. Oh, no. No.

1:36:13

No. No. You could still -- Oh,

1:36:15

you still have authenticator. Or you

1:36:17

could -- Yes. -- be here. Oh,

1:36:19

Yes. Yes. Yes. No. I did that on Twitter too. You you had to have

1:36:21

SMS to enable to FA on Twitter.

1:36:23

But once you'd

1:36:26

set up a key, or an authenticator, you could then disable it. So

1:36:28

you're saying you do the same on

1:36:30

PayPal? Yes. Although there is no UI for

1:36:32

doing Oh, interesting. Just You need to

1:36:34

con turn off You need to contact

1:36:36

them. Yeah. You have to contact them

1:36:38

and say, please unconfirm the phone number. And that makes sense. Right? because

1:36:40

you the phone number I don't have

1:36:42

with somebody else. Right. Right. Yeah. Anyway,

1:36:46

it was Simon who originally noticed and

1:36:49

communicated that it was always possible

1:36:51

to cause PayPal to send

1:36:53

an SMS code for account you know

1:36:55

slash password recovery. Yeah. However, I should note someone

1:36:57

else sent me a note and I'm I

1:37:00

apologize to

1:37:02

that person for let letting us slip letting his name slip.

1:37:04

But he sent me a

1:37:06

note that if if users

1:37:08

if users said set

1:37:10

up their own personal account

1:37:11

recovery recovery questions. You know, those

1:37:13

like, you know, who was your

1:37:16

favorite high

1:37:18

school teacher? and what was the name of your first dog or

1:37:20

whatever. If you set those

1:37:23

up, they cannot be

1:37:26

bypassed. So that's another solution.

1:37:28

Deliberately choose impossible to

1:37:30

guess no matter how well

1:37:32

someone knows you, account recovery

1:37:36

questions, and assuming that

1:37:38

that information is correctly provided,

1:37:40

then you'll be

1:37:42

safe from hijacking because nobody else will know what it was that that you up. It's

1:37:44

just TWiT more passwords

1:37:47

basically. Yeah. Yeah. Yeah. Okay.

1:37:51

Finally, I mentioned last week that I

1:37:54

that I thought, spin right's

1:37:56

new AHCI driver

1:37:58

was not

1:37:59

working correctly. Leo

1:38:01

was wrong about that. It

1:38:03

was working correctly. It was the location in my code where I was

1:38:07

taking the hash of

1:38:10

pin rights results that was causing a false positive detection. So I found

1:38:12

and fixed that and

1:38:14

made some other final improvements.

1:38:18

Then as planned, I up I I

1:38:21

updated GRC's server to get it

1:38:23

ready to manage all

1:38:26

subsequent downloads of the prerelease testing versions of

1:38:28

Spinrite that will be forthcoming.

1:38:30

That work is

1:38:31

finished and the server has

1:38:34

been restarted and is now standing

1:38:36

by to make SPINRAID available.

1:38:38

I have one final feature to add, which came up

1:38:41

about ten

1:38:44

days ago. SPINRAID six

1:38:46

one has four levels or degrees of its operation.

1:38:51

The first level never performs any writing

1:38:53

to a drive under any circumstances. It's strictly

1:38:56

read only. I'm

1:38:59

not sure why, but it always seemed like it

1:39:01

ought to offer that, so

1:39:03

it always has.

1:39:06

The second level is allowed to perform data recovery.

1:39:08

So it will selectively

1:39:10

rewrite only those regions

1:39:14

of the media that are in need of repair.

1:39:16

Level three goes further.

1:39:18

Since refreshing any drives

1:39:21

data is generally

1:39:23

good for it, And

1:39:25

that's because latent and

1:39:27

evolving soft errors are completely hidden by all modern drives. Level

1:39:32

three always rewrite the drives

1:39:34

data as it's moving through the drive. And level four

1:39:39

goes even further. writing in writing

1:39:41

inverted data, reading it back to verify it,

1:39:44

then rewriting

1:39:47

the original data. and reading it back

1:39:50

to make sure that it was written correctly. Okay. I mentioned this

1:39:53

because there

1:39:56

are three classes of drives that

1:39:58

I refer to as being right hostile

1:40:00

and should

1:40:03

only be used under Spin

1:40:06

Rights first two, read mostly levels. Those drives are

1:40:12

SSDs, whose media we

1:40:14

know is incrementally fatigued by writing to it.

1:40:16

Hybrid drives, which

1:40:19

incorporate an SSD, on

1:40:22

their front end to serve

1:40:25

as a nonvolatile cash

1:40:27

and SMR

1:40:28

drives. where

1:40:29

SMR stands for shingled magnetic

1:40:31

recording. Shingling,

1:40:33

exactly like

1:40:36

it sounds, refers

1:40:38

to the deliberate overlapping of

1:40:41

adjacent

1:40:41

tracks in order

1:40:44

to push track

1:40:46

density to insane levels. If you

1:40:49

picture a shingled roof, you cannot

1:40:51

change

1:40:51

an embedded shingle without pulling

1:40:54

up the

1:40:55

shingle above it. and

1:40:57

then the shingle above that one, and the shingle above that one and

1:40:59

so on. The same is true for

1:41:01

SMR drives, which makes writing

1:41:04

to them something

1:41:07

you wanna do as little as possible. As I

1:41:09

mentioned, this issue just came

1:41:12

up in Spin Rights News

1:41:14

Group discussion a couple of weeks

1:41:16

ago. Since I

1:41:17

want SpinRite to continue doing everything possible for its user, in this

1:41:19

case, warning them if

1:41:22

they are about to

1:41:24

perform a

1:41:26

level three or four scan on any drive, which should not be written to needlessly.

1:41:33

I need to

1:41:34

be able to detect that, but I didn't own any hybrid or SMR drives. So I

1:41:36

immediately tracked some

1:41:39

down on eBay and

1:41:42

those four drives have all arrived.

1:41:44

The last two just came in yesterday's mail.

1:41:46

So after today's podcast, I'll be adding detection

1:41:50

of those drive technologies to spin right

1:41:52

so that it can take

1:41:55

responsibility for warning its

1:41:57

users if they're about to do something

1:41:59

that they probably don't wanna do. And then with that little Leo with that last

1:42:01

bit of technology in place, as far

1:42:03

as I know, SPINRAID

1:42:07

six one will be ready to start its

1:42:09

final stage of prerelease

1:42:12

testing. And as

1:42:14

for that, I'm absolutely certain there

1:42:16

are things I missed, things I just

1:42:18

can't see because I'm their author, but that's why we test. What

1:42:22

I am confident of is that at this point, so much testing has

1:42:24

already been done by far the

1:42:26

bulk of the work, that there

1:42:29

are no show

1:42:32

stoppers remaining it should be a matter of

1:42:34

cleaning up debris. So by next week's podcast, it will have been under test

1:42:36

for I'm hoping

1:42:38

that this is a Thanksgiving.

1:42:41

present for our testers,

1:42:43

so I should have a good calibration on where we stand. Nice.

1:42:48

Incidentally, Project

1:42:51

Hale Mary is the book of the month

1:42:53

for Stacey's Book Club in January. If

1:42:55

you have read it or wanna read it,

1:42:57

that's a good book to read. And discussion.

1:42:59

We all read it and loved it. It was a great book.

1:43:01

And if you can listen to the audiobook,

1:43:03

there's some features the audiobook

1:43:05

has that the written

1:43:08

page cannot that makes it kind of fun

1:43:10

too. Anyway, it's good either way. Would you like to come back and talk about

1:43:14

why Peep in just a bit late. Alright. First,

1:43:17

a word from our

1:43:19

sponsor SecureWorks SecureWorks

1:43:23

is a leader You probably

1:43:26

I'm sure you know the name

1:43:28

in cyber security. They build solutions for security

1:43:30

experts, buy security experts, they offer superior threat

1:43:32

detection, and

1:43:34

rapid incident response all while making sure customers, and then you'll like this, are never locked into

1:43:39

a single vendor. SecureWorks

1:43:42

offers an open extended detection and response platform tajus, XDR, extended detection

1:43:44

and response. It's now

1:43:47

time to get it.

1:43:49

TWiT you've

1:43:51

been thinking about it, this is it.

1:43:53

This year cybercrime will cost

1:43:55

the world its

1:43:58

estimated seven trillion

1:43:59

dollars with a t

1:44:01

by twenty twenty five ten

1:44:03

point five trillion. Last

1:44:06

year, ransomware totaled twenty billion dollars

1:44:08

in damages that we know. A tax

1:44:11

occurred every eleven seconds. It's estimated

1:44:13

ten years later, ransomware will

1:44:15

cost two hundred six fifty five billion dollars

1:44:17

a year and strike every two seconds. And I think that's

1:44:19

the optimistic. That's the optimistic

1:44:23

guess. Make sure your organization is not the next victim. You

1:44:25

don't wanna be in those stats. With

1:44:27

SecureWorks TAGES, XDR.

1:44:31

SecureWorks TAGES, superior detection you need, identifying

1:44:33

more than get this four

1:44:35

hundred seventy billion security

1:44:39

events a day. A day. They've got

1:44:41

their they've got their feelings

1:44:43

out everywhere. They

1:44:46

prioritize true positive alerts. They eliminate all that alert

1:44:49

noise, which means you're gonna focus

1:44:51

on the real threats. But

1:44:53

it's important that you get that

1:44:55

intelligence right, that you know what's going on out there. In addition, Tejas

1:44:57

offers unmatched response with automated

1:44:59

response actions. And that

1:45:01

way, because they're automated,

1:45:04

you eliminate the threats before the

1:45:06

damage is ever done. Fast response is key in all of this. With SecureWorks'

1:45:08

TAGES managed XDR,

1:45:11

you can easily leverage those

1:45:13

great SecureWorks experts to investigate and respond to threats on your behalf.

1:45:15

This helps you cut dwell times,

1:45:18

decrease operational, burden, reduce cost,

1:45:23

And with TAGES and SecureWorks managed XDR,

1:45:25

you've got twenty four seven

1:45:27

by three hundred sixty

1:45:30

five day a year coverage.

1:45:32

What does that mean? Well, if you experience

1:45:34

a Christmas day security event or half your team is

1:45:36

out sick, you don't have to

1:45:39

worry. You can trust SecureWorks. is

1:45:42

behind you. Of course, these days, everybody's suffering a lack of a dearth of security

1:45:44

talent. It's hard to

1:45:47

find those people. Right? Don't

1:45:50

worry SecureWorks acts as an extension of

1:45:52

your security team on day

1:45:55

alleviating cybersecurity talent gaps, which

1:45:58

means you, can customize the approach and the coverage

1:46:00

level you need and never

1:46:02

be caught. You know, I don't

1:46:04

wanna say with your pants down,

1:46:06

but You don't wanna be surprised. Let's put it that way. What

1:46:08

happens if you've already found an intruder in your system?

1:46:10

I want you to write this. Get a piece

1:46:14

of paper. Write this down. one eight hundred

1:46:16

breached. Even if you're not a customer,

1:46:18

one eight hundred breached that number connects

1:46:20

you with the

1:46:22

SecureWorks emergency incident response

1:46:25

They can provide you with immediate assistance any

1:46:27

time of the day or night, and they can respond to and remediate a

1:46:29

possible cyber incident

1:46:31

or data breach. one

1:46:34

eight hundred breach. Put it put it put

1:46:36

it in your in your wallet. Put it

1:46:38

on a posted note. At SecureWorks, you

1:46:40

can learn more about the ways today's threat

1:46:42

environment is evolving. the risks it presents to your organization. They've got

1:46:45

case studies. They've got reports from

1:46:47

their very very good counter

1:46:50

threat unit. and a whole lot more. Here's what you do right

1:46:52

now. SecureWorks dot com slash twit. Go there right

1:46:55

now, get a free trial of TAEUS

1:46:57

XDR. No words I

1:46:59

can use to try Leo really give you

1:47:02

that full scope of what they do. It's kind of amazing. SecureWorks dot com

1:47:04

slash TWiT. Get

1:47:07

that free trial. SecureWorks. defending

1:47:10

every corner of cyberspace secureworks dot com slash

1:47:13

TWiT.

1:47:15

to it Thank

1:47:17

you. SecureWorks for supporting everything we do

1:47:19

here at security. Now now whatever it is,

1:47:21

I wanna know what

1:47:23

is y WiPeep. Little

1:47:26

-- Okay. -- peep.

1:47:28

Little white peep. So

1:47:30

imagine our technology that

1:47:32

allows someone walking past a

1:47:35

multistory building or a drone flyby to accurately

1:47:37

locate and pinpoint within

1:47:39

that building or any

1:47:41

other similar space closed

1:47:44

or open. with

1:47:46

a positional accuracy of about

1:47:48

a meter, the location of every

1:47:50

WiFi device, such as security cameras,

1:47:54

and locks and switches, and

1:47:56

anything else on WiFi. That capability,

1:47:58

which, you know, jumps off

1:48:00

the pages of science fiction movie

1:48:03

scripts is not only here now,

1:48:05

but it costs about twenty

1:48:07

dollars. The two researchers who

1:48:09

figured out how to make

1:48:11

this WiFi mapping technology real, named at

1:48:14

They presented their research during the recent ACM

1:48:18

mobile com twenty two,

1:48:20

which was held last month

1:48:22

in October in Sydney, Australia. Here's

1:48:25

how they describe what

1:48:27

they accomplished. They said, we present

1:48:30

Waipipeep, a new location revealing privacy

1:48:32

attack on

1:48:35

non cooperative Wi Fi

1:48:37

devices. Wide exploits

1:48:40

loopholes in the 802

1:48:43

dot eleven to elicit responses from WiFi

1:48:45

devices on a network that we do

1:48:48

not have

1:48:51

access to. It then uses a novel time of

1:48:53

flight measurement scheme to locate

1:48:55

these devices. Why Peep

1:48:58

works without any hardware? or

1:49:01

software modifications on target devices and without requiring access

1:49:03

to the physical space that

1:49:07

they're deployed within. Therefore,

1:49:09

a

1:49:09

pedestrian or a drone that carries a y peep

1:49:12

device can

1:49:15

estimate the location of

1:49:18

every WiFi device in a building. Our Y Peep design costs twenty dollars

1:49:20

and

1:49:21

weighs less

1:49:23

than ten grams. We

1:49:26

deploy it on a lightweight drone and

1:49:28

show that a drone flying over

1:49:30

a house can estimate the

1:49:33

location of WiFi devices

1:49:35

across multiple to meter level

1:49:37

accuracy. Finally, we investigate different

1:49:40

mitigation techniques to

1:49:42

secure future WiFi devices

1:49:44

against such

1:49:46

attacks. Okay. So, you know, this this

1:49:49

has never been

1:49:52

done before. that

1:49:54

the key components here are the the non cooperative nature and the

1:49:57

fact that this

1:49:59

that this

1:50:01

is from this is

1:50:03

being done by a probe which

1:50:06

is not on the WiFi network.

1:50:08

So so they set

1:50:10

this up and and

1:50:12

frame the problem explaining

1:50:14

the problems they encountered and

1:50:16

how each such problem was

1:50:19

solved. They they said, we live in an era of WiFi connected

1:50:22

TVs, refrigerator, security cameras,

1:50:26

and smart sensors. We carry

1:50:28

personal devices like smartwatches,

1:50:30

smartphones, tablets, and laptops. Due

1:50:32

to the deep penetration, of

1:50:35

WiFi devices into our lives.

1:50:37

Location privacy of these

1:50:40

devices is an important and

1:50:42

challenging objective. Imagine a drone that flies over your home

1:50:45

and detects the location of all

1:50:47

your WiFi devices. It could

1:50:49

infer the location of

1:50:52

home occupants. security cameras,

1:50:54

and even home intrusion sensors. A burglar could use this information to

1:50:59

locate valuable items like laptops

1:51:01

and identify ideal opportunities when people are either not at

1:51:03

home or away from

1:51:06

a specific area. For

1:51:08

example, everyone is

1:51:10

in the basement by tracking

1:51:13

their smartphones or smartwatches. The

1:51:15

promise of pervasive connectivity has

1:51:18

been to merge our physical

1:51:20

and digital worlds, but the leakage

1:51:22

of such location information brings arguably

1:51:26

the worst aspect of the digital

1:51:28

world pervasive tracking into

1:51:30

the

1:51:30

physical world. In

1:51:33

this paper, we show that

1:51:35

there are fundamental aspects of the WiFi

1:51:37

i triple E802

1:51:39

dot eleven protocol that

1:51:41

leak such location information to a potential attacker. We

1:51:44

demonstrate that it

1:51:46

is possible to reveal accurate

1:51:50

location of all WiFi devices in an indoor environment,

1:51:56

a, noncooperatively, without

1:51:59

any coordination with WiFi devices

1:52:01

or the access

1:52:03

points.

1:52:03

b, instantaneously,

1:52:04

without

1:52:06

waiting for devices to organically transmit

1:52:09

packets and see

1:52:12

surreptitiously. Without any

1:52:15

complex infrastructure, deployment in the in the surrounding. Our goal

1:52:17

is to expose the security

1:52:19

and privacy vulnerabilities of

1:52:23

the 802 dot eleven WiFi protocol by

1:52:25

demonstrating a first of

1:52:27

its kind non

1:52:31

cooperative localization capability. We hope that

1:52:34

our work will inform the design of next generation

1:52:36

protocols.

1:52:39

So I said, we note that been much work past

1:52:42

work in WiFi based positioning.

1:52:44

However, such

1:52:45

past work does

1:52:48

not enable non cooperative,

1:52:50

certitious localization of WiFi devices. First,

1:52:53

most

1:52:54

of this work relies

1:52:56

on cooperation from end devices. For example,

1:52:58

the client needs to switch channels or

1:53:01

physically move or share

1:53:03

inertial sensor data.

1:53:06

Second, state of the

1:53:08

art techniques such as array

1:53:10

track rely on antenna arrays

1:53:13

with multiple antennas that

1:53:15

are typically bulky and cannot be easily

1:53:17

carried by a person or a small drone. Deploying multiple such

1:53:19

antenna arrays near

1:53:23

a target building makes the attack less

1:53:25

practical and easier to detect. And I don't know if they said, but way more expensive,

1:53:28

obviously. Third,

1:53:32

RSSI based, and remember that's

1:53:34

received signal strength indicator, RSSI based

1:53:37

techniques rely

1:53:40

on fingerprinting or trained models

1:53:42

that require physical access to the target space. Finally, most

1:53:44

of these needed

1:53:47

most of these need

1:53:49

client devices to continuously transmit WiFi packets or

1:53:51

share their received WiFi

1:53:54

packets by installing an

1:53:56

application. an

1:53:59

access we cannot assume for

1:54:01

such privacy revealing

1:54:03

mechanisms. So they say

1:54:05

we present WiPeep peep.

1:54:07

a system that

1:54:07

is quick, accurate, and performs

1:54:10

non cooperative localization.

1:54:13

It does not require

1:54:15

any access to devices or the network access points.

1:54:18

It does not even need the

1:54:20

attacker to

1:54:23

connect to the same WiFi network. In our attack,

1:54:25

the attacker, a lightweight drone, or

1:54:27

a pedestrian, passes

1:54:30

by the house. carrying a small Wi Fi capable device

1:54:32

and estimates the location of

1:54:34

all Wi Fi devices in

1:54:38

the target environment. We exploit the design of the

1:54:40

802 dot eleven protocol

1:54:42

to first generate WiFi traffic

1:54:46

from non cooperative clients.

1:54:49

Then use a novel time

1:54:51

of flight based technique to locate these devices, why solves the

1:54:56

following challenges. Okay? The

1:54:58

first challenge, generate WiFi traffic without cooperation. They

1:55:02

explain, we must, a,

1:55:04

we must a Identify

1:55:06

all devices in the

1:55:09

network quickly at the start of

1:55:11

the attack. And b, generate

1:55:14

WiFi traffic continuously from such devices to perform location

1:55:20

estimation. A simple solution to

1:55:22

identifying devices is to passively wait for WiFi devices

1:55:25

to transmit

1:55:28

a packet. This approach is problematic

1:55:30

because it requires the attacker to linger around for a long

1:55:35

time. Instead, we

1:55:36

exploit the 802 dot

1:55:38

eleven power saving mechanism, which is available in all

1:55:41

802 dot

1:55:43

eleven standards from eleven a

1:55:46

and b up through eleven

1:55:49

ax. By injecting

1:55:51

a fake beacon Imitating the

1:55:54

access point that tells all connected Wi Fi devices to

1:55:57

contact the

1:56:00

access point to receive

1:56:02

buffered packets. This beacon elicits a response from

1:56:04

all devices on the

1:56:07

target Wi Fi network. Once

1:56:11

we've identified all devices, we use

1:56:13

targeted packets to each of

1:56:15

these devices. To

1:56:18

perform Time of flight measurements on these

1:56:20

devices, the attacker requires

1:56:22

exchanging packets directly with

1:56:25

target devices. Therefore, natural traffic from

1:56:27

a target device cannot be

1:56:30

used. Recent work has

1:56:32

shown that 802

1:56:35

dot eleven devices always respond to

1:56:38

packets with an act. Even when the packets emerge outside

1:56:40

the WiFi network

1:56:43

and are unencrypted, or

1:56:47

incorrectly encrypted. We

1:56:49

use this flaw to

1:56:52

perform time of flight

1:56:54

measurements to any target The challenge

1:56:56

in using WiFi is that WiFi devices are in

1:56:58

the sleep mode most of the time

1:57:01

and their radios

1:57:04

turned off. We have designed a technique

1:57:06

that allows an attacker to keep the radio of target devices on

1:57:08

during the attack so

1:57:10

that they keep sending acts.

1:57:14

Okay. So basically, what

1:57:16

these guys did was

1:57:18

was to recognize there was

1:57:20

a way to to

1:57:23

to to after learning about the beacon in a

1:57:25

in a residence or

1:57:27

a corporate facility

1:57:30

or wherever, to simulate a broadcast from

1:57:32

the beacon, which will induce

1:57:35

all WiFi devices to

1:57:37

respond. when they respond, they're gonna

1:57:40

get each device's MAC address.

1:57:42

That then allows them to

1:57:45

individually target those devices selectively.

1:57:48

And in real time, you know

1:57:50

so basically, they they they

1:57:53

get an instant inventory and

1:57:55

then they switch into an active tracking

1:57:57

mode where they are their

1:57:59

their spewing out

1:58:02

packets, measuring round trip

1:58:04

time, what they what they which

1:58:06

they call time of flight in order to determine their instantaneous distance

1:58:10

Gibson are away from

1:58:13

each of the devices. And

1:58:16

of course, as they move, all

1:58:17

of those various vectors are changing length

1:58:20

and by changing

1:58:23

their path, they're able to infer where

1:58:25

the device must be in

1:58:27

order for its vector to

1:58:29

have changed as it did

1:58:32

over time. So then they

1:58:34

explain the second problem they had was localization

1:58:36

in the

1:58:39

face of noisy what

1:58:42

they call SIFS, which is short for short interframe

1:58:44

space. So they explain in

1:58:46

802 dot eleven, acts

1:58:51

are sent at a fixed interval

1:58:53

after receiving a data

1:58:55

packet. This interval

1:58:58

is called short interframe space or

1:59:00

SIFS, as illustrated in the figure

1:59:02

that they have in their notes.

1:59:05

They said, Wipe measures the round trip time between

1:59:07

a packet transmission and an accuracy

1:59:12

reception and subtracts the

1:59:14

SIFS. This allows YPF to estimate the time of flight

1:59:16

and hence the distance between

1:59:18

the attacker and the target device.

1:59:22

Unfortunately, our experience our

1:59:25

experiments reveal that even

1:59:27

though the WiFi

1:59:31

protocol mandates SIFS to be ten microseconds. In practice,

1:59:33

this delay can vary

1:59:36

from eight to

1:59:38

thirteen microseconds. Such errors can

1:59:42

randomize the location estimation process. We build a new algorithm

1:59:44

to correct for such

1:59:46

variations in time of flight

1:59:52

estimates. And finally, dealing

1:59:54

with multi path effects.

1:59:56

They explained that the

1:59:59

time of

1:59:59

flight measurements are error prone because

2:00:01

multiple copies of a signal arrive at

2:00:04

the receiver

2:00:07

from multiple paths. you know, reflection

2:00:09

of signals within an environment. They said the strongest path may not

2:00:12

necessarily be

2:00:15

the direct path. Since the

2:00:17

attacker is far away and obstructed from the target, this problem is

2:00:19

further exacerbated. Indeed,

2:00:23

our measurements reveal that

2:00:26

WiPeep Peep's individual time of flight measurements are error prone for this reason.

2:00:28

To counter this

2:00:31

challenge, we take

2:00:32

what

2:00:35

they call the wisdom of the crowd approach.

2:00:37

Even though each measurement

2:00:39

is noisy, WiPeep

2:00:41

Peep involves quick packet act

2:00:44

sequences at the millisecond level. So

2:00:46

they're doing, you know, thousands per

2:00:50

second. Therefore, we can collect hundreds of

2:00:52

measurements as the attacker flies

2:00:54

by or walks by the

2:00:58

target. we exploit the spatial diversity

2:01:00

of these measurements to get an

2:01:02

accurate position estimation of our targets. So

2:01:05

though you

2:01:06

know, that's a brilliant and completely workable

2:01:08

solution. Individual measurements are noisy,

2:01:11

but the truth can be

2:01:13

found by sorting through thousands

2:01:15

of measurements made over time from different positions.

2:01:18

And then they talk about their implementation. They

2:01:20

said, we've

2:01:23

implemented our design on an

2:01:26

ultralight DJI mini

2:01:30

two drone. You probably have one Leo off

2:01:33

the shelf. Well, I have them

2:01:35

in three, but okay. Actually, it's

2:01:37

something I can finally Leo can

2:01:39

do with Yeah. Yeah. There's a picture in there

2:01:41

of paper, but it's kinda cool. Yeah. Yeah. It's neat. Yeah.

2:01:44

Sort of like stuck

2:01:46

on the front of Yeah. I don't know how willing to fly with that on there,

2:01:48

but I guess it's not too heavy.

2:01:50

They managed to do it. Yeah. Anyway,

2:01:54

they said using off the shelf, ESP thirty two,

2:01:56

and ESP eighty two sixty

2:01:58

six Wi Fi modules. Our

2:02:01

hardware weighs ten grams and cost less than

2:02:03

twenty dollars. It could be deployed on

2:02:06

lightweight drones or carried by a

2:02:08

person. Our evaluations in

2:02:11

a real environment shows that

2:02:14

YPIP finds the location of target devices in an 802 dot eleven ax Wi

2:02:19

Fi six network on three

2:02:21

different three different floors of a house with

2:02:23

a median error of

2:02:26

one point two meters in

2:02:30

around two minutes. The contributions of this paper are.

2:02:33

We present a

2:02:36

new way for

2:02:38

802 dot eleven protocol

2:02:40

features to perform time of flight based

2:02:42

positioning of WiFi devices without having any control

2:02:46

over target devices.

2:02:48

We find that many devices

2:02:50

deviate from the standard time

2:02:53

for SIFS which creates a challenge for

2:02:56

localization. We design a localization

2:02:58

technique that finds a target

2:03:00

device without knowing the exact

2:03:03

SFSIFS used by the device. We present

2:03:05

a solution for future

2:03:07

WiFi chipsets that

2:03:11

allows authenticated devices to perform

2:03:14

localization while disabling non cooperative attacks.

2:03:19

though So Consider these facts,

2:03:21

which they then enumerate.

2:03:24

The WiPeep

2:03:28

peep attacks Work with any WiFi device

2:03:30

without instrumentation, in other words, without any application or firmware

2:03:35

level changes. It does not need physical access to the enclosed

2:03:37

physical space and does not

2:03:39

need to break the

2:03:42

encryption of the WiFi network.

2:03:44

Once the target MAC addresses

2:03:46

obtained, the target device doesn't even be connected to WiFi. Due

2:03:51

to the ease of attack, why Peep has many privacy

2:03:54

and security limitations they write.

2:03:57

We list some examples some example implications

2:04:00

below. In these scenarios, we

2:04:02

assume that it is common

2:04:04

for a person to carry

2:04:06

a WiFi capable device such as a smartphone or a smartwatch.

2:04:09

Also note that the type

2:04:11

of device, iPhone versus

2:04:14

smart sensors can be identified through various

2:04:16

means like the vendor specific

2:04:18

information in the MAC address.

2:04:22

Okay. So and

2:04:24

they give us four examples.

2:04:26

One one impacting security. An

2:04:29

attacker contract the location of

2:04:31

security guards inside sensitive buildings. For example, banks if

2:04:34

they carry a smartphone

2:04:37

or a smartwatch and notice that this

2:04:39

is real time. So moving targets are fine. They

2:04:44

will get time feedback as

2:04:46

things move within the area that they're surveilling. A privacy

2:04:51

implication, an eavesdropper, can fly

2:04:53

a drone over a hotel to find the number and types of

2:04:56

rooms currently

2:05:00

occupied. This could be done by

2:05:02

a rival hotel trying to find detailed information of how target business is

2:05:07

performing. that belong to a room such

2:05:09

as smart TVs can be filtered

2:05:12

based on

2:05:14

Mac addresses. If other devices such as tablets and

2:05:16

laptops are found in a room,

2:05:18

it can be considered occupied, and this

2:05:20

could be done in the middle of

2:05:23

the night when most guests are in

2:05:25

their rooms, or a privacy security implication. If the

2:05:27

MAC address of a device that belongs

2:05:29

to a person of

2:05:31

interest is known, White

2:05:33

Peep contract that person individually in a crowd -- Oh,

2:05:35

excuse me. -- or

2:05:39

inside a building, like

2:05:42

a shopping center or an airport, even

2:05:45

when their device is

2:05:47

not connected to any

2:05:50

WiFi network. So this is so you could tail

2:05:52

somebody with one of these in

2:05:54

your pocket. Yep. That's interesting. Yeah.

2:05:58

Security, why could be used by burglars

2:06:00

to find out the occupancy

2:06:02

status of specific parts of a

2:06:04

building. For example, the burglar

2:06:06

can find out all the people are

2:06:08

on the second floor and the basement

2:06:10

is empty. White peep can also be used for positive use cases.

2:06:12

And I like

2:06:15

this, for example, In a hostage situation,

2:06:17

the police can fly a drone over the building

2:06:19

to find out where

2:06:22

the hostages are kept because

2:06:25

many hostages might have smart devices on them, and they

2:06:27

would be collected together in a

2:06:29

dense group and

2:06:32

not moving. TWiT might also

2:06:34

be possible to track the attackers as well. Okay. Then, anyway, through the balance of their paper,

2:06:37

which

2:06:40

is lengthy, they proceed to

2:06:42

deal with every aspect of their system and present its solution.

2:06:45

So

2:06:47

my point is, The method

2:06:49

to do this today is now in the public domain. So

2:06:51

anyone who wants to do it and

2:06:53

has the skill set to

2:06:55

replicate their work can.

2:06:58

You know,

2:06:59

I could do that.

2:07:00

Many of our listeners could do that.

2:07:02

And I would not be surprised if

2:07:05

we didn't eventually see an off the

2:07:07

shelf turnkey white peep mapping

2:07:10

system that would allow

2:07:12

anyone with only a few

2:07:14

dollars to spare, to obtain this potentially

2:07:16

powerful remote Wi Fi

2:07:18

mapping capability, very much the

2:07:21

way script kitties are using scripts

2:07:23

that they were unable to write.

2:07:25

Until now, we've had a

2:07:27

general sense that the

2:07:29

goings on inside our

2:07:31

homes and offices were at least moderately private.

2:07:33

The idea that someone standing outside in the

2:07:36

middle of

2:07:38

the night could first take a complete inventory of

2:07:41

all WiFi devices within

2:07:43

the area. Noncooperatively, without

2:07:46

connecting to or knowing our

2:07:49

network's password and then

2:07:51

determine the approximate location

2:07:53

of every one of those

2:07:55

devices whether they're upstairs or downstairs, and generally wear,

2:07:57

might not be unsettling to

2:07:59

some people. But there

2:08:02

are likely

2:08:02

some situations and installations

2:08:05

where having such knowledge in

2:08:07

real time could be very valuable to the wrong of people.

2:08:12

The authors

2:08:12

spend some time

2:08:14

near the end of their paper

2:08:16

talking about possible future mitigations. And the

2:08:19

overall outlook there is bleak. Leo

2:08:21

bad news is that since this is a hardware level attack with

2:08:23

all which only

2:08:28

leverages standard WiFi features,

2:08:30

which are implemented in the core WiFi silicon, nothing can

2:08:33

be done in

2:08:35

firmware or software.

2:08:38

all WiFi chips today

2:08:41

will and do respond

2:08:43

to the probe request

2:08:46

packet sent during the use of this technology.

2:08:48

It will take a

2:08:51

future generation of WiFi

2:08:53

chips to deliberately break the

2:08:55

WiFi specification or the spec to

2:08:57

be updated in order to sanction

2:08:59

this by not replying

2:09:02

within a microsecond or

2:09:04

two but by

2:09:06

deliberately randomizing whoa. Excuse me. Deliberately

2:09:08

randomizing

2:09:10

Leo short

2:09:13

interframe space interval,

2:09:15

so that time of

2:09:17

flight information cannot readily

2:09:20

be determined. doing

2:09:22

that will allow WiFi to work while still making location impossible. So

2:09:29

Anyway, that is why Apple randomizes

2:09:31

Mac addresses on its iPhones though. I wonder if that is effective as

2:09:36

a countermeasure. Actually, that's different

2:09:38

than this. This doesn't need MAC addresses. But if you were if I were following you around,

2:09:40

I would need your MAC

2:09:42

address to know it's you. I'm

2:09:46

not saying about the mapping feature. That is true. Yes. Yes. Yes. That is true. addresses

2:09:51

as we know are fixed when

2:09:53

when when the phone is attached to a network -- Right. -- they're they're only

2:09:55

randomized when it's not when it

2:09:58

hasn't joined the network. Once it

2:10:00

has, then

2:10:02

then it uses its actual MAC address. But

2:10:04

you're right. Following you around

2:10:07

the MAC address, I

2:10:10

forgot exactly what the algorithm means. I think they change it every fifteen

2:10:13

minutes. But Yeah.

2:10:15

And I wonder if, you

2:10:17

know, since you know it's

2:10:19

him for fifteen minutes, and then the

2:10:21

MAC address changes. There might be some way to say, ah, yeah. That's he's just

2:10:23

changed his MAC address. I don't know. Not

2:10:25

in the not in the

2:10:28

crowd. Not accrued. You would be

2:10:30

you yeah. Because you would be getting you you would be so first of you be

2:10:33

you'd be

2:10:36

only pinging that. And

2:10:38

then suddenly, there would be no reply. Yeah. Yeah. So if you then go dead and and you You would have

2:10:41

lost it by

2:10:44

that. Right. Yeah. You have to

2:10:46

go back into broadcast mode -- Right. -- in order to get replies from everybody in the neighborhood. I'm less concerned

2:10:49

about somebody mapping

2:10:52

my house. I knew

2:10:54

you wouldn't be from WiFi access points, but the tracking thing is concerning. I think others

2:10:58

I think there's Android

2:11:01

funds that also randomize macros. Well, remember that it's listed.

2:11:03

It's not the it's not the WiFi access point that they're locating. It's

2:11:05

all your security cameras.

2:11:07

Right. Right. And WiFi.

2:11:11

Right. Anything anything WiFi. Yeah.

2:11:13

Yeah. Yeah. Again, less worried

2:11:16

about that.

Unlock more with Podchaser Pro

  • Audience Insights
  • Contact Information
  • Demographics
  • Charts
  • Sponsor History
  • and More!
Pro Features