Latest / Mini Scroll / MINI TECH SCROLL: Tea app data breach, AI model in Vogue, YouTube ban for under-16s & the Internet Archive’s new role

Transcript

MINI TECH SCROLL: Tea app data breach, AI model in Vogue, YouTube ban for under-16s & the Internet Archive’s new role · Mini Scroll

• 0:00
  Hey everyone, welcome to Mini Scroll, the daily Internet
• 0:02
  culture podcast. Bring me the biggest stories
• 0:04
  from social media, the creator economy and the digital space
• 0:07
  every Monday through Thursday and sometimes Fridays.
• 0:10
  My name is Annabelle and I am your Wednesday mini tech Scroll
• 0:13
  host and also the deputy editor of the youth digital media brand
• 0:17
  Centennial World. Welcome back to mini Tech
• 0:20
  Scroll, everyone. Lauren has already spoken about
• 0:23
  this on Monday and Tuesday, but we are rebranding our
• 0:27
  newsletter. If you want to learn more about
• 0:29
  the rebrand, Lawrence spoke about it a little bit more in
• 0:32
  Monday's episode, but we also published an edition about the
• 0:36
  whole rebrand and what to expect, so please check it out,
• 0:39
  it would be much appreciated. We are really excited for what
• 0:43
  comes next with the newsletter. Other than that, Please remember
• 0:45
  to rate, review, and subscribe. And if you want to support
• 0:48
  Centennial in other ways, check out the Buy me a Coffee link in
• 0:51
  the description if you are so inclined.
• 0:54
  OK, let's get into the content for this week.
• 0:56
  The first story today I have seen everywhere and that is all
• 1:00
  all about the T app data breach. So if you're unfamiliar with the
• 1:04
  app, I will just like break it down a little bit for you.
• 1:06
  So T is a women's only app that is essentially meant to help
• 1:10
  women stay safe while dating men.
• 1:13
  It was reportedly launched in 2023 where users could
• 1:17
  essentially post pictures of the men that they match with on
• 1:19
  dating apps or were talking to and get the quote UN quote T on
• 1:23
  them. Users would essentially be able
• 1:25
  to say if they are a green flag or a red flag or any other
• 1:29
  comments they have about their experience with that specific
• 1:32
  person. In the words of the Tea website
• 1:35
  Quote, founder Sean Cook launched Tea after witnessing
• 1:39
  his mother's terrifying experience with online dating
• 1:42
  not only being catfished, but unknowingly engaging with men
• 1:45
  who had criminal records. Realizing that traditional
• 1:48
  dating apps do little to protect women, Sean knew something had
• 1:51
  to change. That's why he built Tea, the
• 1:54
  first ever dating safety platform for women.
• 1:57
  I should note here that many articles that I've read talking
• 2:00
  about the Tea app note that similar kind of things do exist
• 2:04
  on a smaller scale. Many women might turn to
• 2:07
  Facebook groups or other community groups to talk about
• 2:10
  their experience or question the men that they will potentially
• 2:13
  be dating. And honestly, some type of
• 2:15
  crowdsourcing information slash background checking is
• 2:18
  definitely important in this era of modern dating.
• 2:21
  But I will say I do think an app like this can quickly devolve
• 2:25
  into cyberbullying. I never use the app so I can't
• 2:28
  say for sure if that's what was happening on T, but I think it's
• 2:31
  just something to think about. Even though I do think something
• 2:34
  like this is definitely important for the safety of
• 2:36
  women. And obviously I hate that an app
• 2:38
  like this even has to exist in the fast place.
• 2:41
  Like I said, I haven't used the app, but according to the terms
• 2:44
  and conditions on the T website, a user has to create a username
• 2:48
  and include their location, birthdate, photo and photo ID
• 2:53
  when they start using the app. And many women had downloaded
• 2:57
  and used the app where just a couple of days ago it became the
• 3:00
  most downloaded app on the US Apple App Store.
• 3:03
  And while Tea may have started off as an attempt to keep people
• 3:07
  safe, it has ended up doing quite the opposite where TEA has
• 3:11
  found itself in the middle of a pretty bad Dudley.
• 3:13
  At the time of recording, there has reportedly been two
• 3:16
  breaches. So pulling from the TEA website
• 3:20
  once again, it seems that the initial breach was at 6:44 AM
• 3:24
  Pacific Standard Time on Friday, July 25th, where they quote,
• 3:28
  identified unauthorized access to our system.
• 3:32
  It soon became apparent that data from before February 2024
• 3:35
  had been accessed. So according to the statement on
• 3:39
  the T website, they described what had been compromised in the
• 3:44
  breach, where it was quote approximately 72,000 images,
• 3:48
  including approximately 13,000 selfies and photo identification
• 3:53
  submitted by users during the account verification and
• 3:56
  approximately 59,000 images publicly viewable in the app
• 4:00
  from posts, comments and direct messages.
• 4:03
  But soon there were reports that these images had been downloaded
• 4:07
  and were now accessible on 4 Chan.
• 4:09
  So I was reading all about this breach from four O 4 media.
• 4:12
  I would definitely recommend you to check them out if you're
• 4:15
  interested in tech industry, all that kind of stuff.
• 4:18
  But Emmanuel Maybag explained that the hackers gained access
• 4:23
  through an exposed database hosted on Google's mobile app
• 4:26
  development platform Firebase. That being said, according to
• 4:30
  everything I've seen online, this was not not a sophisticated
• 4:33
  hack by any means, and it was seemingly more the fault of the
• 4:36
  platform and the security measures that they did not have
• 4:40
  in place. I've seen experts describe the
• 4:43
  security measures as downright negligent and lazy, so that is
• 4:47
  definitely not positive on behalf of the platform and I'm
• 4:50
  sure there will be legal action coming.
• 4:53
  But earlier this week, T posted another update saying that they
• 4:56
  had recently learnt that some DMS were accessed as part of the
• 4:59
  initial incident. Because of this, they ended up
• 5:02
  taking the DM functionality down.
• 5:04
  In a follow up article for 404 Media, the journalist Emanuel
• 5:08
  Mayberg and Joseph Cox talk about some of the DMS that had
• 5:11
  been exposed, saying that it was possible for the hackers to see
• 5:15
  quote messages between users discussing abortions, cheating
• 5:18
  partners, and phone numbers they sent to one another.
• 5:21
  They also spoke to a researcher who contends that a second
• 5:24
  breach could access data from users who were joined after
• 5:28
  2024, which is obviously the initial cut off date that
• 5:31
  fatigue, the app had said was breached.
• 5:33
  So like I mentioned earlier, a lot of the data and images that
• 5:37
  had been exposed in the breach have ended up on 4 Chan.
• 5:40
  And this is obviously extremely dangerous for many reasons, but
• 5:44
  especially because what has been done with the data is that a
• 5:47
  searchable public map of T users has been created based on the
• 5:52
  IDs that had been accessed. So people can literally see
• 5:55
  information about a person where they live on this interactive
• 5:58
  map. And it is really scary because
• 6:01
  since T started for gaining popularity and virality,
• 6:03
  especially in the past month, it became heavily criticized by
• 6:07
  men's rights activists and groups, where having the
• 6:10
  personal information of many T users is now falling in the
• 6:14
  hands of such men on 4 Chan. There have also been reports
• 6:18
  that users on 4 Chan are ranking the selfies of the users that
• 6:21
  they have come across in times of attractiveness, which we
• 6:25
  hate. From a bigger picture standpoint
• 6:28
  though, we do talk a lot on this podcast about the importance of
• 6:31
  verification and how it is becoming increasingly common
• 6:34
  online as apps introduce age restrictions and verification
• 6:38
  processes to prove that a user is not a bot.
• 6:42
  Verification is definitely important, but the T data breach
• 6:45
  shows how precarious it can be to use photo IDs and selfies
• 6:49
  like this. I think it's particularly
• 6:50
  worrisome in the era of deep fakes where these selfies could
• 6:53
  be used in other videos. Ryan Broderick, who is
• 6:57
  definitely one of our favorite commentators, also covered the
• 7:00
  tea breach in his Garbage day newsletter, and I think he had a
• 7:03
  really, really good point As for what this kind of verification
• 7:07
  means for the future of the Internet and surveillance.
• 7:10
  So I will read to you what he wrote now because he honestly
• 7:13
  said it better than I ever could quote.
• 7:15
  Monitoring one kind of user means monitoring everyone else.
• 7:19
  Proving that a user is a woman poses the same problem with the
• 7:22
  additional thorniness of defining what a woman is.
• 7:26
  A quandary T didn't survive long enough to reckon with.
• 7:29
  But the lesson from all of this is that there is no simple
• 7:31
  solution here. Instead we have found ourselves
• 7:34
  facing 2 choices. Fight for the chaotic open
• 7:37
  Internet that allows anonymity and all the good and bad that
• 7:41
  comes with it. Or continue to slide into an
• 7:43
  Internet that feels safer but surveils our every move and will
• 7:47
  inevitably censor what we see and do, supported by massive
• 7:50
  databases of our most embarrassing and sensitive data.
• 7:54
  So I thought that was really interesting and I'd be intrigued
• 7:57
  to hear which choice out of the ones that Ryan Broderick lays
• 8:01
  out, you would prefer. My next story that I have today
• 8:04
  though, is another one that I've been seeing quite a lot of
• 8:07
  commentary about, and that is how an AI model was included in
• 8:10
  Vogue. Not last week, but I'm pretty
• 8:12
  sure the week before I covered a story about how an AI influencer
• 8:16
  was going viral for her post that looked like she was at
• 8:19
  Wimbledon. In the episode, I spoke about a
• 8:21
  number of things. I did go in a bit of a rant, I
• 8:23
  will say, but I did touch on what it means for the rise of AI
• 8:27
  influences and what it represents for the creator
• 8:30
  economy. But obviously we've seen a
• 8:32
  rising number of AI influences and just AI content more broadly
• 8:36
  online over the past few years. But I will say, I feel like it
• 8:39
  has been mostly relegated to the digital world in mainstream
• 8:44
  media. We haven't really seen a massive
• 8:46
  uptick in AI and visual content yet.
• 8:49
  I feel like, correct me if I'm wrong, maybe it's just what I'm
• 8:51
  seeing, but behind the scenes, I'm sure media professionals are
• 8:55
  using it in script, writing, special effects and whatever.
• 8:58
  But I do feel like we haven't really seen the rise of slop in
• 9:01
  the same way as we have online in more mainstream forms of
• 9:06
  media and entertainment. But it seems like the seal has
• 9:09
  officially been broken. In the August edition of Vogue,
• 9:12
  a picture of an AI generated model was included in a guest
• 9:15
  campaign. And there is obviously so much
• 9:18
  to discuss here. But first let me kind of explain
• 9:21
  what the picture was. Obviously if you are watching, I
• 9:25
  will insert it in the screen, but if you're listening, not to
• 9:28
  worry. So the AI generated women had
• 9:31
  blonde hair, light eyes and the quote UN quote perfect model
• 9:34
  body. There was a small disclaimer
• 9:36
  indicating that the image was AI generated but it was very faint.
• 9:41
  So the images were created by a london-based AI marketing
• 9:44
  agency. The founders of the agency told
• 9:47
  the BBC that they managed to get the campaign after a guest Co
• 9:51
  founder reached out to them in DM.
• 9:52
  I thought this was interesting because it does seem like these
• 9:55
  calls are coming from the top. And I know this is a little bit
• 9:57
  different, but I actually included an article in our old
• 10:01
  newsletter once again, please subscribe a couple weeks ago
• 10:05
  about how AI is infiltrating the media landscape.
• 10:09
  It was published once again by 4 or 4 Media.
• 10:11
  They're really getting a lot of airtime today.
• 10:12
  But the journalist noted how ironic it is that the higher ups
• 10:16
  and executives of so many of these media companies in the
• 10:18
  United States are pushing for the inclusion of AI as you know,
• 10:22
  the quote UN quote shiny new thing, despite the fact that it
• 10:26
  is really hurting the industry, limiting creativity and critical
• 10:30
  thinking in the process. I feel like while AI models like
• 10:33
  this could be a form of creativity and art, you could
• 10:37
  make a similar argument. It also to me feels particularly
• 10:40
  since start to talk about all of this in the shadow of the Sydney
• 10:43
  Sweeney cross American Eagle campaign.
• 10:46
  Lauren covered this yesterday where she spoke about the
• 10:48
  messaging of the campaign and how it kind of feeds into light
• 10:51
  supremacy, the turn to conservatism culturally and also
• 10:55
  eugenics. And the fact that this AI
• 10:58
  generated model is also a woman who fits neatly into Eurocentric
• 11:03
  beauty standards suggests that even when there are endless
• 11:05
  possibilities to create whatever type of image woman that you
• 11:09
  want in a campaign like this one, the fashion brands once
• 11:12
  again fall into this mold. And to be honest, I don't know,
• 11:16
  call me optimistic, but I really was hoping for something
• 11:19
  different from the fashion industry.
• 11:21
  But I don't know, I probably shouldn't have got my hopes up
• 11:24
  on that one, especially because I feel like we have moved so far
• 11:28
  away from a beauty standard that kind of valued original beauty
• 11:32
  towards something where beauty equals perfection.
• 11:35
  Like the beauty standard today involves getting work done on
• 11:38
  your face or your body in one way or another.
• 11:40
  And now we have these AI models that are not even real that can
• 11:44
  further uphold these beauty standards.
• 11:46
  I can't, I cannot separate though from the wrongdoings of
• 11:49
  the industry at large and obviously the values it clearly
• 11:53
  upholds. These developments also just go
• 11:55
  to show how capitalism and the pursuit of profit literally
• 11:58
  shapes everything. Fashion at its core, I feel like
• 12:01
  should be about creativity like I was kind of getting at before,
• 12:06
  but now it's essentially just an ends to a mean to make money.
• 12:10
  And sure, you could say that it's always been like that and I
• 12:12
  probably agree, but it's just so much more blatant now.
• 12:15
  Also trying to make money by showing clothes on a person that
• 12:20
  doesn't exist? How does that even make sense?
• 12:23
  I don't understand. Once again I feel like I've gone
• 12:26
  on another rant, but I think another thing worth discussing
• 12:30
  is that if we have the rising popularity of AI models, there
• 12:33
  is obviously issues about cutting work for real people.
• 12:37
  Which is something similar to what I discussed when I was
• 12:40
  talking about the AI influences. But also with these AI models
• 12:44
  and influences, they obviously pull from real life people.
• 12:48
  Like we know AI is always pulling from something, it's not
• 12:50
  entirely original. So when they created this model,
• 12:54
  whose likeness are they even pulling from?
• 12:57
  Should this passing be referenced?
• 12:58
  There are so many ethical questions here.
• 13:01
  Like I've seen so many people commenting that the AI model in
• 13:04
  this campaign looks like Kate Hudson.
• 13:06
  But yes, that is my take on that.
• 13:08
  I feel like I just blacked out. I don't even remember anything I
• 13:11
  said, so hopefully that all made sense.
• 13:14
  OK, to finish off today's episode, I actually have two
• 13:17
  shorter and hopefully quick stories.
• 13:19
  The first one is an update on something I covered a couple
• 13:23
  weeks ago and that is about YouTube and the social media ban
• 13:26
  for under sixteens here in Australia.
• 13:28
  As a refresher, last year legislation was passed through
• 13:31
  Australian parliament that would set a minimum age for social
• 13:34
  media use, including platforms like Snapchat, TikTok, Facebook,
• 13:38
  Instagram and X YouTube was not initially included in that list
• 13:42
  of platforms, but maybe around a month ago there seemed to be a
• 13:46
  push that YouTube should be included in the quote UN quote
• 13:49
  ban. And yesterday it broke that
• 13:52
  YouTube would in fact be included.
• 13:55
  This is obviously a big move because while there is no
• 13:58
  denying that YouTube has problems, especially with
• 14:01
  sending viewers and young people down rabbit holes, it is
• 14:04
  obviously a major educational tool.
• 14:07
  Having said that though, I have read that YouTube Kids is
• 14:11
  expected to be exempt from the ban, so there is still some type
• 14:14
  of access there. Nevertheless, the Australian
• 14:16
  government's decision comes days after Google, which is obviously
• 14:19
  the parent company of YouTube, threatened legal action if the
• 14:22
  government decided to include the platform in the ban despite
• 14:26
  originally exempting it. Obviously, this isn't really
• 14:29
  surprising because it makes total sense that YouTube would
• 14:31
  work really hard to stay off the ban list because it allows them
• 14:35
  to keep more users on the platform, watch more videos, and
• 14:38
  of course, make more money. Anyway, that is just a quick
• 14:41
  update for you on that. My last story for today though
• 14:44
  is that the Internet Archive is now being used as an archive for
• 14:48
  the government in the United States.
• 14:49
  We actually did a whole deep dive on the Internet Archive, so
• 14:53
  definitely check that out if you were interested.
• 14:55
  I found that one quite interesting, but if you haven't
• 14:58
  watched that episode I will quickly explain what the
• 15:00
  Internet Archive is. So it's a tool that essentially
• 15:03
  let's Internet users view links and digital materials that have
• 15:07
  been deleted or altered, which ultimately make them challenging
• 15:10
  to find online. It also runs the Wayback
• 15:13
  Machine, which I'm sure many of you have used over the years,
• 15:16
  but you can basically just drop a link in and it will take you
• 15:20
  to older versions of said website anyway.
• 15:23
  Recently, the Californian Senator Alex Padilla granted the
• 15:27
  Internet Archive federal depository status, which allows
• 15:31
  it to more easily access and share government documents with
• 15:34
  the public. In a statement, Padilla said
• 15:37
  quote the archives Digital first approach makes it the perfect
• 15:40
  fit for a modern federal depository library, expanding
• 15:43
  access to federal government publications amid an
• 15:46
  increasingly digital landscape. The Internet Archive has broken
• 15:49
  down countless barriers to accessing information.
• 15:52
  And it is my honor to provide this designation to help further
• 15:55
  their mission of providing universal access to knowledge.
• 15:58
  I thought this was pretty interesting because the Internet
• 16:01
  Archive has not been without its problems over the years.
• 16:04
  For example, it's been hit by a series of lawsuits and cyber
• 16:08
  security issues. Recently, the organization has
• 16:11
  specifically faced legal issues around intellectual property and
• 16:14
  copyright. At one point, the Internet
• 16:16
  Archive was sued by four major book publishers who argued that
• 16:20
  scanning physical books for digital access violated be Right
• 16:23
  law in the United States. So yeah, I'm wondering if this
• 16:26
  sort of validation will end up helping the Internet Archive in
• 16:31
  the long term, especially because I do believe having some
• 16:34
  sort of archive for digital materials is really important
• 16:37
  because that is the main way we communicate these days.
• 16:40
  Mashable, the publication also made an interesting point around
• 16:43
  this whole situation, noting that, quote, the status is
• 16:47
  particularly notable as the Trump administration has
• 16:50
  systemically removed information from federal websites under new,
• 16:53
  quote, anti woke executive orders.
• 16:56
  So in some ways, this designation is actually
• 16:59
  completely at odds with this kind of agenda from the
• 17:02
  government. But yeah, I'm intrigued to see
• 17:04
  what happens here. OK, those are all my stories I
• 17:07
  have for today. I feel like that was a bit of a
• 17:09
  long one, or maybe I'm just tired, who knows.
• 17:12
  But Lauren will speak to you tomorrow and I will chat to you
• 17:17
  next week. Once again, Please remember to
• 17:19
  sign up to our newsletter. The link is in the description.
• 17:23
  OK, thanks for listening. Bye.