Latest / Parsing the Truth: One Byte at a Time / Will Emojis Keep You Safe?

Transcript

Will Emojis Keep You Safe? · Parsing the Truth: One Byte at a Time

0:01
Join former FBI Senior Forensic Examiner's Becky Passmore and
0:05
Stacey Eldridge as they dig into the world of digital forensics
0:08
one bite at a time. Now leading their own consulting
0:11
firms and educating the next generation of cyber sleuths,
0:14
these badass babes of digital forensics bring sharp insight,
0:18
real world case experience, and a dose of humor to the bits and
0:21
bites that build today's investigations.
0:23
From deep dives into business e-mail compromise and ransomware
0:26
attacks, Becky's specialty to cracking the complexities of
0:29
iPhone forensics, Stacy's obsession, this podcast delivers
0:33
insider knowledge for professionals, students, and
0:36
true crime techies alike. If you've ever wondered what
0:39
really happens when the digital trail is all that's left, TuneIn
0:42
and start parsing the truth. Welcome back to Parsing the
0:53
Truth One Bite at a Time, where we cut through digital noise to
0:56
reveal the facts. I'm your Co host, Stacey
0:59
Eldridge, and as always, I'm joined by my Co host, Becky
1:02
Passmore. And in today's episode, we're
1:05
talking about emojis, AI, online safety and privacy.
1:10
Woo Hoo. Hey everyone.
1:12
Today we're diving into something a lot of parents do
1:16
without thinking twice. And Stacey, I see this often,
1:19
you know, being in some social media sites, I see people
1:23
posting photos of their kids online.
1:26
And then let me just cover my child's face with this cute
1:30
little heart emoji sticker. Yeah.
1:33
And it's almost the default move these days, right?
1:36
I'm just going to throw a heart over their face.
1:39
That'll keep my kids safe. Nobody will know what they look
1:41
like. In this episode, we want to get
1:43
to the truth, especially with the spike in articles that came
1:46
out about this topic at the beginning of the school year.
1:50
So we're focusing on will that :) or that emoji or that
1:54
sticker. Whatever you slap on your kids
1:56
picture online, is it going to keep your child's identity safe
2:01
like you think it is? Right, you know, and this is
2:03
really a great topic, something that we need to be talking
2:07
about, especially with, you know, the holidays quickly
2:10
approaching and coming on to us as fast as it can.
2:13
So, you know, let's start with a little bit of the psychology
2:17
behind it. Most parents don't post their
2:21
children online with bad intentions.
2:24
That's not the reason, right? They're proud.
2:27
They want to share those milestones, first day of school,
2:31
sports, holidays, all the things.
2:33
And so they post online with those intentions.
2:37
Yep, and covering bring that face, your little kids face with
2:40
that emoji feels like a privacy solution.
2:43
You can still show off the outfit, the Halloween costume,
2:48
the event, whatever that cute moment is and you think, OK, no
2:53
one can see their face, so I'm good.
2:55
It's what the Independent called security theater in their
2:59
article about this topic. It's when you're feeling safer,
3:02
but you haven't actually reduced all the risks and you haven't
3:06
actually reduced all the risks. And it has become trendy.
3:10
You see more and more people doing it.
3:12
You see the influencers, family, bloggers, even celebrities are
3:15
doing it. And maybe the celebrities kind
3:18
of started it most of all, right?
3:20
But there's all sorts of emojis going on, the faces, hearts,
3:23
flowers, and there's even pixelated filters.
3:25
But what a lot of people don't realize is that the face is just
3:29
one data point in that picture. That's right.
3:32
And it's not just about what's covered, it's about everything
3:36
else that's still visible in that picture.
3:40
So let's talk about the three main articles that really digged
3:46
into this stuff. And the first one is from that
3:49
independent like you mentioned, that was back in June of 2025.
3:53
And this cyber security expert Lisa Ventura claimed the
3:57
covering of face with an emoji doesn't remove other identifying
4:02
information. Things like the background, the
4:04
school logos, the furniture, when a backyard.
4:08
I mean, even like if I had a backyard fence in my backyard,
4:12
all those details that can still help piece together some
4:17
information where that photo has been taken.
4:20
Yep. Exactly.
4:22
So even though the face is hidden, the digital bread crumbs
4:26
are still out there. Now, another article for ABC
4:29
News Australia points out that these emojis can block some of
4:34
the automated facial recognition systems.
4:36
But they said that's only part of the problem because what's
4:40
happening is just because the algorithm can't match the face
4:45
to the human and some AIS with theoretically and maybe some
4:51
still can recognize the person from context clues.
4:55
So they're saying I might not have your face in this picture
5:00
to reconstruct it, but I've gotten it from a different
5:04
picture that's been uploaded online because Lord knows what
5:07
happens as we're uploading all this stuff into AI.
5:11
Who can see it? Are the IAIS talking to each
5:13
other on the back end, you know, and passing all this information
5:17
together? Like how secure is any of this
5:19
stuff? And it's just, you know, as
5:21
we've said before, once you post something online, it's out there
5:24
forever. So there could be that there,
5:26
All right, There could be that. So you could be able to take
5:30
your face from a different picture and reconstruct it on to
5:34
another face with the help of AI.
5:36
But then a news article from the Star went a step further and
5:39
they warned that the new AI tools can actually remove emojis
5:44
or stickers quite easily and that they can reconstruct faces.
5:48
Or perhaps the worst thing is invent plausible new faces on
5:53
your pictures. Listen, I was playing with that
5:57
today and I uploaded a photo to add some caption to it and there
6:03
was five faces in that photo that were average people, like
6:10
probably in their 40s, five people.
6:13
It added the caption I wanted it to add it to.
6:15
I was just playing around with it and it made the faces look
6:19
like they were 70s and 80 year olds.
6:22
So it took them from being like 40 year olds.
6:26
That's the wrong direction. Right.
6:28
It aged them 40 years ahead. It was astonishing.
6:33
I was like, oh, OK, well, we'll see how that works.
6:35
Yeah, so. And that's what's so unsettling
6:38
with this. These tools aren't hypothetical.
6:41
They're already online. Yeah.
6:43
They're free or low cost. I think you and I have talked
6:46
about the different versions and the free cost version versus the
6:51
paid version. Not much more is shared or given
6:54
out, but it's still using these platforms.
6:58
And if we use them for object removal or photo repair instead
7:02
of going out there and using, you know, like Adobe Photo
7:06
Picture to edit it, deciding to upload that picture to AI and
7:11
have the same object removal or photo repair completed.
7:15
And it's the same technology that's used to erase watermarks
7:19
or remove photo bombers. And then it's also called Image
7:24
in painting, and we're going to break it down for you.
7:27
So imagine you have a photo of your child with a big heart
7:31
emoji covering their face. Yeah.
7:35
And when you post it, that heart is flattened like a pancake into
7:41
the image. It becomes part of those pixels.
7:43
So you think there's no face behind it anymore, but lo and
7:47
behold. I mean and, and that's true, the
7:50
original pictures under the heart are gone.
7:54
And we wanted to dig into this a little more with a photographer
7:57
expert. So we spoke with Bonnie at
7:59
Genesis Photography here in Lincoln, NE to find out as a
8:04
Photoshop expert what could she do with these heart covered
8:07
faces. Bonnie is just a whiz when it
8:10
comes to a Photoshop and I'm not just talking like oh she can get
8:14
rid of my blemishes and my fine lines.
8:17
She does these imagination scenes so she will pit her
8:23
subject and just like out of these world situations.
8:28
Very beautiful and it's looks all very realistic.
8:32
It does not look fake at all. So she knows she has that touch.
8:35
She knows how to do all the fancy things and photo to make
8:40
an impossible situation look real.
8:43
So first her approach was she downloaded some random pictures
8:47
from Facebook that had hearts covering the faces.
8:50
And 1st she tried to figure out if these pictures that she
8:53
downloaded had any layers in them that could be removed and
8:57
they did not. And that was no surprise to her.
9:00
And if you've not worked with Photoshop and I know just enough
9:04
to be dangerous, when you work in Photoshop, you can make your
9:07
projects and then you can create all of these different layers.
9:10
And so that's how you kind of work on the different things to
9:13
like shadow the one thing out or dim this and brightness, you're
9:16
using all these different layers to the project.
9:19
And that saved in a proprietary Photoshop file format.
9:23
But then if you wanted to give it to me or put it on the
9:27
Internet, that sort of thing, you've got to put it into a JPG
9:31
format so other people can open it without having Photoshop.
9:35
And when you do that, you have to do it in Photoshop.
9:37
It basically flattens everything down like Becky was saying that
9:41
heart emojis flattened down and it's compressing everything into
9:45
that JPEG format. So there's no more layers.
9:48
It's compressed. And the more you save it and
9:50
work with it, the more what's called noisy it gets.
9:55
I'm not the photo expert so I couldn't tell you like what
9:58
photo looks more noisy. But the more times you re say
10:01
stuff and and work with it and that JPEG, it does potentially
10:05
lose some of that quality over the time because you're just
10:08
recompressing it over and over again.
10:11
So she was not surprised that there were no layers.
10:14
And this is kind of sort of what I expected as well.
10:17
Yeah, and so Bonnie also said there was another great
10:20
Photoshop tool called Lightroom and that would not even open the
10:25
photos saved from Facebook from excuse me, from Facebook.
10:29
Because of that, Bonnie was only able to work with the photos in
10:34
Photoshop, and Photoshop made it all weird colors and lines.
10:39
She then prompted the AI generative fill in Photoshop to
10:45
remove the icons and reveal their faces.
10:49
And what she got was pretty spectacular.
10:52
I spectacularly awful Stacey, why don't you show them what she
10:58
was able to come up with cuz it's really kind of creepy.
11:02
Yeah, I guess it's timely with Halloween coming up right upon
11:07
us. So it kind of fits, fits that.
11:10
So let me get ready to screen share here.
11:13
See, OK, so this is the picture that down.
11:18
This is a picture that Bonnie randomly downloaded from
11:21
Facebook. So it's got the kiddos with your
11:24
typical heart emojis covering your face.
11:27
And so we can't see their faces. But like we talked about, you
11:32
can see some identifying information in here.
11:36
There's some sort of name for something kids, Doctor White
11:40
official. So I mean there is some
11:42
information in here that you could read and the same back
11:46
here. Medals around their neck.
11:48
They have things on their shirts.
11:52
Might be like a contest that might say something, but.
11:55
Yeah, there. There you go.
11:56
Kids run. DeWitt, Oliphant.
11:58
Right. And so someone could just look
12:01
up where that run occurred, and then you can see back there it
12:04
says Schoenberger. Schoenbeck I don't know, right,
12:11
You know, so I mean, this is some of the identifying
12:14
information that Woo we were talking about.
12:17
That's in the in still in the picture.
12:19
So yes, you can exactly see what your kiddos look like.
12:22
But like, here's this race that they just took place in.
12:25
Let's get back to what she did next.
12:29
So she opened it up in Photoshop and this is what it looked like.
12:33
See how it's all the lines and it's got this weird purplish and
12:37
green color. And so the AI tool in Photoshop
12:41
is generative AI Phil and she said remove the heart and reveal
12:45
the child's face. And this is what it came up
12:49
with. Super, super creepy.
12:52
So half of the photo is still got the weird lines running
12:57
through it and that was the side with presumably a boy because of
13:02
the short buzz cut. But now the boy in the AI
13:06
generated photo has like longer hair and a Bob and looks like he
13:11
has some purple lipstick on. That's right, very big lips now.
13:15
And then on the other side of the picture, there's no no more
13:19
pixelated lines running through it.
13:21
However, this girl has bushy crazy hair now.
13:26
Before was it? Was it in a ponytail?
13:30
I already forgot. Dad, no, it was, it was not.
13:33
It was still a little on the Bob side, but yeah, yeah.
13:38
Not as bad as the boys, Not as yeah.
13:41
But it did make it hair bigger, a little bit bushier.
13:45
But the really creepy thing is what it did to her face, it made
13:50
her look super. Oh, it reminds me kind of of a
13:55
witch. But it's like doing what you
13:57
said, right? It added like 40 years on it did
14:00
to the people in your picture. Clearly this young girl has been
14:04
aged up to, I don't know, 6080 years old and and turned into a
14:10
witch. So that was definitely kind of
14:14
creepy. And then she tried it again.
14:18
And so in this one, the boy has his buzz cut haircut back.
14:24
However, his his face is really kind of no features.
14:28
He's still got the lines right. Almost right.
14:31
Yeah, he's got like a very pointy chin.
14:34
He's got kind of a mouth. And then like, these black,
14:38
dark, dark, dead eyes. Yeah.
14:40
Yeah. It's like, OK, all I'm going to
14:43
do this time is I'm going to give you eyeballs, nose and a
14:46
mouth, right. No depth or texture or anything
14:49
like that. But then on the girl, the hair
14:54
is pretty much back to normal, but her face now looks like a
14:59
plastic doll, Right? Face.
15:03
With. Classically smooth, yeah.
15:04
I don't know why they added glasses, but she has glasses
15:07
now. The eyes are a bright white and
15:11
you got like the dark black pupils in them.
15:14
And then she does have a lips like you.
15:19
I guess you would expect like when you're seeing like a doll.
15:21
Like it kind of reminds me of a Barbie face, just with the
15:25
petite nose and the petite mouth and all that sort of stuff.
15:28
But it clearly looks very plasticky to me and super fake
15:32
in this picture. So I mean, obviously this is not
15:35
a good example of what Photoshop, or maybe it is a
15:39
great example of what Photoshop can do and it's not good.
15:44
But Photoshop is not the only tool out there.
15:48
And this is probably, I mean, maybe someone who's up to no
15:52
good would be using Photoshop, but maybe they're going to be
15:56
using some of those other online tools out there to try to put
16:00
the faces into the pictures, right.
16:02
So I got a celebrity picture. It's Halloween.
16:06
It's coming up. So we've got, is it Channel 7
16:09
news Ron, Ron Burgundy and the Anchorman.
16:14
It's his, his new, his news team is here in this picture.
16:18
And so the kiddos and their costumes have pumpkins covering
16:22
up their faces. And so I thought, OK, I'm going
16:25
to throw this into a new AI tool.
16:29
What was the name of the tool I used?
16:31
It was AI ES because this was really the only one that I could
16:36
find online that I could do for free.
16:39
A lot of them that I found you had to buy credits and you
16:43
couldn't just buy like a credit. You had to buy a pack of
16:45
credits. So they're they're going to get
16:48
your money. A lot of them, but I picked this
16:51
one and I thought this one was kind of fun what it came up
16:54
with. So a little a little bit better
16:57
than what a Photoshop was able to do with the pictures.
17:03
But in front of from what I've read the and from what I've
17:09
read, what AI does is it kind of takes its surroundings to fill
17:14
in the gaps. And so I think you can really
17:17
see that here in this picture. And I'm going to zoom in just a
17:21
touch if you're watching. So for the boy, and I think this
17:26
one was Paul Rudd's character in the movie.
17:29
I don't remember his name, but because he's next to Ron
17:33
Burgundy, so now he's got a mustache and it kind of gave him
17:37
some chops on the side, kind of where the pumpkin was covering
17:41
his face. Right.
17:44
I would say also, you know, it's still got some of that orangey
17:47
pumpkin on the edges, so it couldn't get rid of absolutely
17:51
everything. But it makes me think, OK, Yep,
17:53
this person has a mustache. I'm going to give him a
17:55
mustache. Right, right.
17:57
And then the other guy, the other kid in the blue suit over
18:04
here, I was trying to figure out what the heck was going on with
18:06
his eyes. I feel like they kind of made
18:09
him. A puppet, you know, like they
18:12
take like a puppet. Yeah, yeah, cuz it's very
18:18
smooth, solid face. It looks like maybe he has like
18:21
a little bit of a stash going on, but he's got like a weird
18:26
colored face. And then like there's some weird
18:29
stuff going on. Like I was like, are they
18:31
putting glasses over his face? But it's definitely a weird
18:37
face. And I was trying to figure out,
18:38
OK, who whose eyeballs or face is that coming from around
18:43
there? Not sure what's going on.
18:44
I think it tried to put glasses on him that that's what I think.
18:47
I think it's a because here it looks a little reflectively, but
18:52
then this is maybe the what do you call the the the thing that
18:57
keeps it on your ear, your. Earpiece.
19:00
Yeah. OK.
19:01
Earpiece or whatever the side piece of the glass is like sits
19:04
on your ear. That one side looks like it bows
19:06
out a little bit. I don't know it just it just
19:08
looks weird. It's giving it a shot.
19:11
It's certainly better than what Photoshop was coming up with
19:14
because these kind of look like actual faces with some
19:18
personality. But then the problem could be,
19:21
you know, people inventing faces to put on your, your body.
19:25
So. So that can be troublesome.
19:28
Yeah. And so I'm going to go ahead and
19:31
stop screen sharing. So, yeah, and then and then on
19:34
top of what Stacey was showing in those pictures, which
19:37
obviously was very disturbing, right.
19:39
Yeah, yeah. We got to talk about what we see
19:42
from a digital forensics perspective when we view the
19:46
photos in hex. Yes.
19:50
But first we're going to, we would use, but first we would
19:53
use the exit tool to see that there is virtually no metadata
19:57
left in the photo. Social media sites strip all of
20:01
the data off of the off these days.
20:04
So now you should be careful where and how you share because
20:07
on Telegram you can post a picture as a file and then
20:12
whoever downloads it will have the original photo complete with
20:17
the metadata. So.
20:19
Do you have these photos where you can show them in hex?
20:24
I could pull them up in hex. Do we want to do that or do we
20:27
want to just show them what EXIF data?
20:29
Yeah, let's just see what. Yeah, let's just see what EXIF
20:31
tool produced. And are you going to pull up
20:34
both of them or are you just going to pull up the one?
20:36
I will do both of them. OK cool so this lost my mouse.
20:42
OK, so this is the first one. Yeah, you might want to zoom in
20:46
a little bit. It's fine.
20:49
It's fine. I'll zoom in.
20:50
I just. OK, I'm making view at the top.
20:53
Where at the top view 3 tabs over.
20:56
Let me talk to you about how to use a tool.
21:02
Yeah, that looks good. OK, we'll get this out there
21:07
that zoomed on both. OK, OK.
21:09
Do you want to talk through it so that and then just tell me
21:11
when you're done and we'll switch.
21:12
So this is the first one that that Bonnie did.
21:15
Yeah. So just a header and no JPEG
21:17
with JPEG to extract, right. So, so we're just talking about
21:20
the header information here. We could also, you know, see
21:25
that the saves it was saved multiple times, right.
21:30
So there's a lot of, there's some date, some data here.
21:35
You have some file modification, access creation, you have a zone
21:41
identifier. Do you want to talk about what
21:42
that means? Zone identifier means what it
21:46
was downloaded from the Internet, right?
21:49
Some, some of our artifacts populate that zone identifier
21:54
when it's downloaded. And this tells us, this EXIF
21:58
tool tells us that that zone identifier exists, which then I
22:03
would go and try to find the artifacts showing and proving
22:06
that it was downloaded from the Internet.
22:10
It gives us the file permissions, looks like read,
22:12
write. So there's no, you know, special
22:16
permissions there. You got the file type, you've
22:19
got the size of it, the image width and height of the image,
22:25
you have some color components and the resolution.
22:29
So it gives us the image size, the megapixels, all the details.
22:34
Not a whole lot. Right, right.
22:36
There's, you know, no identifying information in here
22:39
at all. It doesn't even say that it came
22:41
from Facebook, and it certainly doesn't give GPS coordinates,
22:45
your camera, anything like that. That's right.
22:48
And this photo that Bonnie got. From from Facebook.
22:51
Yep. And then the other photo that
22:53
you downloaded actually. Yep, and this one was from
22:56
Instagram. Yep.
22:58
And did you nameit5.jpg? I didn't, no, it was, it was, it
23:02
was something else and I renamed it.
23:04
Yep. So.
23:05
So it shows the name you gave it, not the name that was there
23:09
before. Again, zone identifier exists.
23:13
It shows the file size of it. It shows the file modification,
23:17
access, creation date. All done today because.
23:21
That's when I downloaded it. That's right.
23:23
And then JPEG it gives us a current IPTC IPTC digest.
23:30
Not really sure what that is, but we could look that up in
23:34
EXIF tool instructions. And then it goes on and it gives
23:39
us some special instructions. Again, the image size, not a
23:43
whole lot of details. Right.
23:44
But this one has slightly more information than the Facebook
23:48
one did, but not a whole lot. And again, we're still missing
23:51
any sort of identifying information, GPS coordinates,
23:55
things of that nature. So from a metadata perspective,
23:59
in terms of privacy, this is good because people are not
24:03
going to get information about you from any photos that you've
24:09
posted, unlock Facebook and Instagram, things like that.
24:13
So that stuff's all going to going to be gone.
24:17
But you could see here just from looking at this too, that the
24:19
quality of these pictures are not great either.
24:22
These are not like printable photo quality photos either.
24:27
Yeah, yeah. So this is great.
24:28
Thank you for sharing that, Stacy.
24:30
Let's dive back into what some of these articles were talking
24:34
about in the terms of tool. Yep.
24:37
So here's where AI changes things with image.
24:40
In painting, the system analyzes all the surrounding pixels, the
24:45
skin tone, the lighting, the hair, the coloring.
24:48
That's predicting what it's and it's predicting what is supposed
24:52
to be behind that covered area. And so it's the same process
24:57
used in restoration tools or apps that, say, remove unwanted
25:02
objects. So what you would do with the
25:04
tool, you highlight the sticker and then AI fills it in.
25:08
And sometimes it's shockingly accurate or sometimes it's
25:13
complete creepy garbage, right? But tools like art, smart, real,
25:19
mine, and even some Photoshop plugins can make a reconstructed
25:23
face that looks 90% believable. And other times it's a complete
25:28
flop, as we've shown you today. And I'm sure like the Photoshop
25:31
pros, if they have them, because, you know, they do
25:33
things like how you have like a big group of people in the
25:37
picture. You're trying to get 20 people
25:39
to look and smile and all at the same time.
25:43
And then like you finally got the shop except for one kids.
25:46
So they do the thing where they take like the head from the
25:48
other photo and put it on to this photo.
25:51
So I'm sure there's some way to do that much easier with
25:55
Photoshop and AI and and those sorts of things.
25:58
But remember, we're talking about like strangers on the
26:02
Internet downloading stuff and so maybe they have been stalking
26:06
you and your family for a while and they've just been collecting
26:11
pictures over. Time.
26:13
Yep, that's exactly right. And that's what the Star,
26:16
another publication, was warning us about warning about.
26:21
Even if it's not your child's actual face, the AI can create
26:27
one that looks realistic enough to misuse it for.
26:31
Profiles are even worse, right? So that cute heart might hide a
26:37
real face from your Instagram followers, but it doesn't stop
26:41
someone with malicious intent or a little technical skill from
26:46
generating one, right? And even if the face can't be
26:50
restored, there's still all that contextual information in the
26:55
background. The same background, the same
26:58
bedroom, same vacation spot, those things can easily be tied
27:04
back to that same child. And it's just from running
27:10
searches on the Internet, you know, a lot of us, I will see
27:14
something really cute like a purse or a pair of shoes or
27:18
something like that. I'm like, I wonder where I could
27:20
buy that. So I'll take a photo of it.
27:23
Then I'll throw it into Google Lens and I'll tell it to give me
27:27
the details. Well, people can do the same
27:29
thing. Give me any details about this
27:32
picture. And quite honestly, they.
27:34
Might wear this. Correct.
27:36
Yes, yes. So might find some more details
27:40
just by running it through through a Google lens, by taking
27:44
a picture or again through AI. Yep, Yep.
27:47
So let's continue to talk about some digital footprints.
27:51
Every time you post a picture, even with a sticker, you're
27:54
leaving little clues. Think about captions like first
27:58
day of third grade or off to soccer at Lincoln Park.
28:03
And then when you combine that information with time stamps,
28:05
suddenly someone is going to be able to track your child's
28:09
routine. And it's not just strangers.
28:11
There's all sorts of people or companies in the background
28:15
doing stuff. There's data brokers, facial
28:17
recognition systems and even platforms themselves where you
28:21
are posting your pictures that can use this information to
28:25
build profiles, especially if you are posting these things
28:30
publicly. Yeah.
28:31
The ABC's. The ABC News article even
28:35
mentions that posting repeatedly overtime creates a pattern, and
28:41
quite honestly that that's what a lot of predators or people
28:46
that are stalking aren't looking for is patterns.
28:48
Even if faith, even if the faces aren't always hidden.
28:53
AI doesn't need one clear photo. It learns from repetition.
28:57
And when you post on social media you also give the platform
29:02
rights to your content. Many platforms terms of service
29:06
say they can use, reproduce, or distribute images for their
29:12
purpose purposes. Once it's up, it's out of your
29:14
control. That's something most parents
29:17
forget about. Even if you delete the post
29:20
later, it might already be archived, scraped, or shared
29:24
with anyone, Right? Exactly.
29:26
And remember, all these companies are doing backups of
29:29
their stuff because, you know, God forbid Facebook go down,
29:33
what would we do? So there's got to be backups.
29:36
There's so they could restore, so the worst happen and all of
29:39
that sort of stuff. So there's definitely multiple
29:42
cops, puppies and their stuff being replicated across
29:44
different servers for the purposes of speed.
29:47
So it's not like you just put one picture and there's one
29:50
picture sitting in one server in one place.
29:53
That's right. Meta headquarters.
29:56
Who knows if it's even physically at Meta headquarters
29:59
because I'm pretty sure it's not.
30:01
No, they don't back their data up at Meta headquarters, do
30:04
they? It's an off site of.
30:06
Course, let's bring in some real world examples.
30:11
I had read about a case where parents had posted a family
30:15
beach photo and the faces were covered with sunglasses emojis.
30:19
But in another picture that was posted in the same series, there
30:23
was a reflection in the sunglasses and that other person
30:27
revealed the child's face in their sunglasses.
30:30
So the children weren't in the picture anymore.
30:32
They had these emojis. But then this other person, you
30:36
know how reflective your sunglasses are?
30:38
You could see the children's face in those sunglasses,
30:41
reflections on. This is classic digital
30:44
forensics material and investigative material.
30:47
Even if you're just looking at regular photos and you're not
30:49
doing it from a digital forensic perspective.
30:52
But we use this all the time. We see it all the time.
30:55
Investigators are looking for reflections, whether it's in
30:58
windows, mirrors, metal surfaces.
31:00
That's why investigators love analyzing photo metadata.
31:04
That's one place to get us started and then the reflections
31:06
to give us more details. You can often reconstruct a
31:09
scene even when and the main subject is blurred or covered.
31:15
Now, this takes me back to a big terrorism case that I worked in
31:19
the FBI. And this person was making
31:23
videos telling people how to turn a remote control, how to
31:28
turn a remote control car into a bomb.
31:32
And so the idea being that you would drive the remote control
31:34
car bomb into wherever and then set the bomb off.
31:38
And so this person thought they were being pretty clever, right,
31:41
because they're recording from like neck down, shoulders down.
31:45
So their face is not in it. It's fairly well zoomed in.
31:49
So I mean, there's not a a ton of stuff in the background.
31:53
It's it's focused on his core body because he's there working
31:57
with the toy car and how to turn it into an explosive device.
32:01
However, on his table was one of those glass table tops.
32:06
And so his face was in the reflection.
32:10
And so we took a bunch of shots of the frames and then they went
32:14
through and did there somebody else did the analysis magic to
32:18
make it a great image that could tie a person to the reflection
32:22
to producing that video. Yeah, that's fantastic.
32:25
Yeah. So let's move from scary to
32:28
practical. What can parents actually do to
32:31
protect their kids privacy? We've got a short checklist
32:35
based on expert recommendations. Number one, host less often.
32:41
Think twice before hitting that upload.
32:43
Does this really need to be online or can it stay in a
32:47
family chat and #2 avoid identifiers.
32:51
That means no visible school logos, no St. games, no unique
32:56
landmarks in the background. Anything that can pinpoint the
33:00
location of where these pictures are being taken should not be
33:05
included in the photo. Yep, it's hard, but you can try
33:08
to crop in as close as possible, sure, and that will help reduce
33:12
your risk. Yeah, for sure. #3 Skip the
33:15
location tags in your social media.
33:18
So turn off Geo posting or turn off Geo tagging when you're
33:22
doing posting. It's fun to share like at the
33:25
park, but it's better to do it after you've left #4 Keep it
33:30
private. Use family only sharing tools
33:33
like shared albums. You can make folders in the
33:36
cloud different services, or just use private messaging apps
33:40
#5 ask relatives to not repost or to post original pictures.
33:46
A lot of grandparents love sharing photos.
33:48
But they not. But they might not realize that
33:52
the privacy settings are wide open, so it's best to have a
33:55
discussion. About.
33:56
Oh yeah, definitely. And number six, don't rely on
33:59
emojis or filters for security. If you want to anonymize a
34:05
photo, use a full blur or a crop instead of a decorative sticker.
34:12
Yeah, a heart is cute, but a crop is safer.
34:15
You can still share a hand holding a trophy or a photo from
34:20
behind without showing the face or location.
34:23
And for parents who really want to post, think about the future
34:28
embarrassment test. Ask yourself, will my child
34:32
thank me for sharing this when they're 16?
34:35
If not make maybe skip it. Problem is, I don't think your
34:38
child will thank you for anything when they're 16.
34:41
Nope, don't think so. Don't think so at all.
34:43
But one of the other things you know, we mentioned, we said, you
34:47
know, keep it to like the family chats, the family channels,
34:50
things like that. But if you are on social media,
34:53
certainly one of the things that can reduce your risk is actually
34:58
knowing the people you are friends with.
35:01
Or are you allowed to follow you on the different social media
35:05
services? So for instance, if you don't
35:09
know who a person is and you want to post lots of pictures of
35:13
your kids on Facebook, then perhaps don't accept that friend
35:18
request. And maybe that means you go
35:20
through and go back through your lists and see who am I friends
35:25
with if you want to keep sharing on Facebook.
35:27
But I think we've given you some ideas of some other places to
35:31
share some of this information with your friends and family.
35:35
That's not quite it's so public and I will say I used to post a
35:40
lot more on Facebook and then I kind of switched to Instagram
35:43
for a while because I had very few people on Instagram.
35:46
But I'm old, so I'm just not really like an Instagram person.
35:50
Facebook's my jam. But now I, I don't really post
35:55
as much on Facebook because neither I, I don't know, I'm
35:58
busy, but like, I really enjoyed looking through my memories on
36:02
Facebook. So that was kind of like while
36:04
I, why I was like, Oh, I got to keep posting some things because
36:07
I want to see the memories, but Google Photos has a memory
36:12
option. So when I go into Google Photos,
36:14
I'll be like, oh, recent memories one year ago today and
36:18
stuff like that. So I'm still getting the same
36:20
effect without sharing everything on Facebook.
36:24
And I certainly did go on Facebook and I've always been
36:28
pretty strict about like who is going to be my friend on
36:31
Facebook? It does need to be somebody I
36:33
know sort of situation. And then for a while, because,
36:37
you know, I was trying to grow a business.
36:38
I had all these people. I was working with and everybody
36:41
wanted to be your friend on Facebook.
36:42
And then so I was trying to manage it by being like, you
36:45
know, giving them the different groups that they can be in on my
36:48
profile. So then I could switch the
36:49
privacy and say who the different people were and that
36:52
sort of situation. And I was like, this is just too
36:55
complicated. I I either know you and my and
36:57
you're my friend. That's right.
37:00
Or you're not and so you're not going to be on on my Facebook
37:03
sort of situation. So that's something everybody
37:05
struggles with. But you know, we're here to just
37:08
give you some tips. So hopefully if you've seen some
37:11
of those headlines like emojis will not save your children or
37:15
the emojis can be removed and everyone will see who your child
37:19
is and what they look like. It's a lot of scary headlines to
37:22
get you to click and to read right now.
37:25
Photoshop I don't think can reliably fill in anything emoji
37:29
wise, but we did see some. It's a little funny, but I mean
37:33
a realistic attempt at filling in a face that had emojis on it.
37:37
Yeah, that's right. Yeah.
37:39
All good tips, all good recommendations.
37:43
You know, we're, we're trying to bring you the truth.
37:45
You know, that's what we talk about all the time, trying to
37:48
bring you the truth. We've seen it from experience
37:51
working in our field that people can be tracked by social media,
37:56
social media posts, social media pictures.
37:58
So that's why we're here and you know, we just want to share this
38:03
very interesting information that we discovered over the last
38:06
couple weeks and communicating with Bonnie and then me just
38:10
playing around with it today and not realizing it was going to
38:13
add 40 years to a picture anyway.
38:16
And maybe most importantly, every post contributes to your
38:20
child's story online. So let's make sure it's one that
38:25
they get to tell someday. So until next time, stay safe,
38:29
stay sharp and as always, keep parsing the truth one by the at
38:33
a time. If you enjoyed this episode,
38:37
make sure to follow us to get notified the next time a new
38:41
episode drops. If you can do Morehead on over
38:44
to parsingthetruth.com and donate by clicking on the donate
38:48
button. Parsing the Truth is a nonprofit
38:50
organization and we are raising money to send a new digital
38:54
Forensic examiner to IASYS training next year in Orlando.
38:58
Thanks.