[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:19.77,Default,,0000,0000,0000,,{\i1}36C3 preroll music{\i0}
Dialogue: 0,0:00:19.77,0:00:25.07,Default,,0000,0000,0000,,Herald: It is my honor to introduce you\Ntoday to Eva and Chris. Eva, she is a
Dialogue: 0,0:00:25.07,0:00:29.44,Default,,0000,0000,0000,,senior researcher at Privacy\NInternational. She works on gender,
Dialogue: 0,0:00:29.44,0:00:34.68,Default,,0000,0000,0000,,economical and social rights and how they\Ninterplay with the right to privacy,
Dialogue: 0,0:00:34.68,0:00:40.43,Default,,0000,0000,0000,,especially in marginalized communities.\NChris, she is the privacy lead at
Dialogue: 0,0:00:40.43,0:00:46.37,Default,,0000,0000,0000,,technology lead at Privacy International.\NAnd his day-to-day job is to expose
Dialogue: 0,0:00:46.37,0:00:51.29,Default,,0000,0000,0000,,company and how they profit from\Nindividuals and specifically today they
Dialogue: 0,0:00:51.29,0:00:59.23,Default,,0000,0000,0000,,will tell us how these companies can even\Nprofit from your menstruations. Thank you.
Dialogue: 0,0:00:59.23,0:01:00.47,Default,,0000,0000,0000,,Chris: Thank you.
Dialogue: 0,0:01:00.47,0:01:05.20,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:01:05.20,0:01:13.86,Default,,0000,0000,0000,,C: Hi, everyone. It's nice to be back at\NCCC. I was at CCC last year. If you heard
Dialogue: 0,0:01:13.86,0:01:18.58,Default,,0000,0000,0000,,my talk from last year, this is going to\Nbe like a slightly vague part 2. And if
Dialogue: 0,0:01:18.58,0:01:21.68,Default,,0000,0000,0000,,you're not, I'm just gonna give you a very\Nbrief recap because there is a
Dialogue: 0,0:01:21.68,0:01:28.38,Default,,0000,0000,0000,,relationship between the two. So, I will\Ngive you a little bit of background about
Dialogue: 0,0:01:28.38,0:01:32.54,Default,,0000,0000,0000,,how this project started. Then we get to a\Nlittle bit about menstruation apps and
Dialogue: 0,0:01:32.54,0:01:38.04,Default,,0000,0000,0000,,what a menstruation app actually is. Let\Nme talk a little bit through some of the
Dialogue: 0,0:01:38.04,0:01:42.25,Default,,0000,0000,0000,,data that these these apps are collecting\Nand talk how we did our research, our
Dialogue: 0,0:01:42.25,0:01:48.39,Default,,0000,0000,0000,,research methodology and then what our\Nfindings are and our conclusions. So last
Dialogue: 0,0:01:48.39,0:01:54.64,Default,,0000,0000,0000,,year, I and a colleague did a project\Naround how Facebook collects data about
Dialogue: 0,0:01:54.64,0:02:03.67,Default,,0000,0000,0000,,users on Android devices using the Android\NFacebook SDK. And this is whether you have
Dialogue: 0,0:02:03.67,0:02:09.54,Default,,0000,0000,0000,,a Facebook account or not. And for that\Nproject, we really looked when you first
Dialogue: 0,0:02:09.54,0:02:13.74,Default,,0000,0000,0000,,opened apps and didn't really have to do\Nvery much interaction with them
Dialogue: 0,0:02:13.74,0:02:23.56,Default,,0000,0000,0000,,particularily, about the automatic sending\Nof data in a post GDPR context. And so we
Dialogue: 0,0:02:23.56,0:02:30.17,Default,,0000,0000,0000,,looked a load of apps for that project,\Nincluding a couple of period trackers. And
Dialogue: 0,0:02:30.17,0:02:36.82,Default,,0000,0000,0000,,that kind of led onto this project because\Nwe were seeing loads of apps, across
Dialogue: 0,0:02:36.82,0:02:42.82,Default,,0000,0000,0000,,different areas of categories. So we\Nthought we'd like hone in a little bit on
Dialogue: 0,0:02:42.82,0:02:48.57,Default,,0000,0000,0000,,period trackers to see what kind of data,\Nbecause they're by far more sensitive than
Dialogue: 0,0:02:48.57,0:02:52.60,Default,,0000,0000,0000,,many of the other apps on there, like you\Nmight consider your music history to be
Dialogue: 0,0:02:52.60,0:03:03.69,Default,,0000,0000,0000,,very sensitive.... {\i1}laughs{\i0} So. Yeah. So,\Njust a quick update on the previous work
Dialogue: 0,0:03:03.69,0:03:11.85,Default,,0000,0000,0000,,from last year. We actually followed up\Nwith all of the companies from that, from
Dialogue: 0,0:03:11.85,0:03:17.45,Default,,0000,0000,0000,,that report. And by the end of like going\Nthrough multiple rounds of response, over
Dialogue: 0,0:03:17.45,0:03:22.41,Default,,0000,0000,0000,,60 percent of them a changed practices\Neither by disabling the Facebook SDK in
Dialogue: 0,0:03:22.41,0:03:30.70,Default,,0000,0000,0000,,their app or by disabling it until you\Ngave consent or removing it entirely. So I
Dialogue: 0,0:03:30.70,0:03:35.69,Default,,0000,0000,0000,,pass over to Eva Blum-Dumontet. She's\Ngoing to talk you through menstruation
Dialogue: 0,0:03:35.69,0:03:38.85,Default,,0000,0000,0000,,apps.\NEva: So I just want to make sure that
Dialogue: 0,0:03:38.85,0:03:42.31,Default,,0000,0000,0000,,we're all on the same page. Although if\Nyou didn't know what a menstruation app is
Dialogue: 0,0:03:42.31,0:03:47.79,Default,,0000,0000,0000,,and you still bothered coming to this\Ntalk, I'm extremely grateful. So how many
Dialogue: 0,0:03:47.79,0:03:53.54,Default,,0000,0000,0000,,of you are are using a menstruation app or\Nhave a partner, who's been using a
Dialogue: 0,0:03:53.54,0:03:58.33,Default,,0000,0000,0000,,menstruation app? Oh my God. Oh, okay. I\Ndidn't expect that. I thought it was going
Dialogue: 0,0:03:58.33,0:04:03.44,Default,,0000,0000,0000,,to be much less. Okay. Well, for the few\Nof you who still might not know what a
Dialogue: 0,0:04:03.44,0:04:07.67,Default,,0000,0000,0000,,menstruation app is, I'm still going to go\Nquickly through what a menstruation app
Dialogue: 0,0:04:07.67,0:04:15.52,Default,,0000,0000,0000,,is. It's the idea of a menstruation app.\NWe also call them period tracker. It's to
Dialogue: 0,0:04:15.52,0:04:21.50,Default,,0000,0000,0000,,have an app that tracks your menstruation\Ncycle. So that they tell you what days
Dialogue: 0,0:04:21.50,0:04:26.72,Default,,0000,0000,0000,,you're most fertile. And you can\Nobviously, if you're using them to try and
Dialogue: 0,0:04:26.72,0:04:32.84,Default,,0000,0000,0000,,get pregnant or if you have, for example,\Na painful period, you can sort of plan
Dialogue: 0,0:04:32.84,0:04:39.66,Default,,0000,0000,0000,,accordingly. So that's essentially the\Nmain 2 reasons users would be would be
Dialogue: 0,0:04:39.66,0:04:48.47,Default,,0000,0000,0000,,looking into using menstruation apps:\Npregnancy, period tracking. Now, how did
Dialogue: 0,0:04:48.47,0:04:53.88,Default,,0000,0000,0000,,this research starts? As Chris said,\Nobviously there was whole research that
Dialogue: 0,0:04:53.88,0:05:01.27,Default,,0000,0000,0000,,had been done by Privacy International\Nlast year on various apps. And as Chris
Dialogue: 0,0:05:01.27,0:05:08.66,Default,,0000,0000,0000,,also already said what I was particularly\Ninterested in was the kind of data that
Dialogue: 0,0:05:08.66,0:05:13.22,Default,,0000,0000,0000,,menstruation apps are collecting, because\Nas we'll explain in this talk, it's really
Dialogue: 0,0:05:13.22,0:05:21.80,Default,,0000,0000,0000,,actually not just limited to menstruation\Ncycle. And so I was interested in seeing
Dialogue: 0,0:05:21.80,0:05:26.82,Default,,0000,0000,0000,,what actually happens to the data when it\Nis being shared. So I should say we're
Dialogue: 0,0:05:26.82,0:05:31.53,Default,,0000,0000,0000,,really standing on the shoulders of giants\Nwhen it comes to this research. There was
Dialogue: 0,0:05:31.53,0:05:35.66,Default,,0000,0000,0000,,previously existing research on\Nmenstruation apps that was done by a
Dialogue: 0,0:05:35.66,0:05:40.93,Default,,0000,0000,0000,,partner organization, Coding Rights in\NBrazil. So they had done research on the
Dialogue: 0,0:05:40.93,0:05:46.69,Default,,0000,0000,0000,,kind of data that was collected by\Nmenstruation apps and the granularity of
Dialogue: 0,0:05:46.69,0:05:52.08,Default,,0000,0000,0000,,this data. Yet, a very interesting thing\Nthat we're looking at was the gender
Dialogue: 0,0:05:52.08,0:05:59.03,Default,,0000,0000,0000,,normativity of those apps. Chris and I\Nhave been looking at, you know, dozens of
Dialogue: 0,0:05:59.03,0:06:03.28,Default,,0000,0000,0000,,these apps and, you know, they have\Nvarious data showing practices, as we'll
Dialogue: 0,0:06:03.28,0:06:07.87,Default,,0000,0000,0000,,explain in the stock. But they have one\Nthing that all of them have in common is
Dialogue: 0,0:06:07.87,0:06:16.15,Default,,0000,0000,0000,,that they are all pink. The other thing is\Nthat they talk to their users as woman.
Dialogue: 0,0:06:16.15,0:06:20.55,Default,,0000,0000,0000,,They, you know, don't want sort of even\Ncompute the fact that maybe not all their
Dialogue: 0,0:06:20.55,0:06:30.28,Default,,0000,0000,0000,,users are woman. So there is a very sort\Nof like narrow perspective of pregnancy
Dialogue: 0,0:06:30.28,0:06:41.02,Default,,0000,0000,0000,,and females' bodies and how does female\Nsexuality function. Now, as I was saying,
Dialogue: 0,0:06:41.02,0:06:45.06,Default,,0000,0000,0000,,when you're using a menstruation app, it's\Nnot just your menstruation cycle that
Dialogue: 0,0:06:45.06,0:06:55.33,Default,,0000,0000,0000,,you're entering. So this is some of the\Nquestions that menstruation apps ask: So
Dialogue: 0,0:06:55.33,0:07:01.09,Default,,0000,0000,0000,,sex; There is a lot about sex that they\Nwant to know? How often, is it protected
Dialogue: 0,0:07:01.09,0:07:08.42,Default,,0000,0000,0000,,or unprotected? Are you smoking? Are you\Ndrinking? Are you partying? How often? We
Dialogue: 0,0:07:08.42,0:07:16.88,Default,,0000,0000,0000,,even had one app that was asking about\Nmasturbation, your sleeping pattern, your
Dialogue: 0,0:07:16.88,0:07:22.93,Default,,0000,0000,0000,,coffee drinking habits. One thing that's\Nreally interesting is that - and we'll
Dialogue: 0,0:07:22.93,0:07:28.91,Default,,0000,0000,0000,,talk a little bit more again about this\Nlater - but there's very strong data
Dialogue: 0,0:07:28.91,0:07:34.07,Default,,0000,0000,0000,,protection laws in Europe called GDPR as\Nmost of you will know. And it says that
Dialogue: 0,0:07:34.07,0:07:38.42,Default,,0000,0000,0000,,only data that's strictly necessary should\Nbe collected. So I'm still unclear what
Dialogue: 0,0:07:38.42,0:07:46.98,Default,,0000,0000,0000,,masturbation has to do with tracking your\Nmenstruation cycle, but... Other thing
Dialogue: 0,0:07:46.98,0:07:56.48,Default,,0000,0000,0000,,that was collected is about your health\Nand the reason health is so important is
Dialogue: 0,0:07:56.48,0:07:59.98,Default,,0000,0000,0000,,also related to data protection laws\Nbecause when you're collecting health
Dialogue: 0,0:07:59.98,0:08:04.73,Default,,0000,0000,0000,,data, you need to show that you're taking\Nan extra step to collect this data because
Dialogue: 0,0:08:04.73,0:08:11.46,Default,,0000,0000,0000,,it's considered sensitive personal data.\NSo extra steps in terms of getting
Dialogue: 0,0:08:11.46,0:08:17.17,Default,,0000,0000,0000,,explicit consent from the users but also\Nthrough steps on behalf of the data
Dialogue: 0,0:08:17.17,0:08:22.06,Default,,0000,0000,0000,,controller, in terms of showing that\Nthey're making extra steps for the
Dialogue: 0,0:08:22.06,0:08:28.79,Default,,0000,0000,0000,,security of this data. So this is the type\Nof question that was asked. There is so
Dialogue: 0,0:08:28.79,0:08:34.56,Default,,0000,0000,0000,,much asked about vaginal discharge and\Nwhat kind of vaginal discharge you get
Dialogue: 0,0:08:34.56,0:08:39.88,Default,,0000,0000,0000,,with all sorts of weird adjectives for\Nthis: "Tiki, creamy". So yeah, they
Dialogue: 0,0:08:39.88,0:08:49.07,Default,,0000,0000,0000,,clearly thought a lot about this. And it\Nis a lot about mood as well. Even, yeah, I
Dialogue: 0,0:08:49.07,0:08:56.19,Default,,0000,0000,0000,,didn't know 'romantic' was a mood but\Napparently it is. And what's interesting
Dialogue: 0,0:08:56.19,0:09:01.90,Default,,0000,0000,0000,,obviously about mood in the context where,\Nyou know, we've seen stories like
Dialogue: 0,0:09:01.90,0:09:07.00,Default,,0000,0000,0000,,Cambridge Analytica, for example. So we\Nknow how much companies, we know how much
Dialogue: 0,0:09:07.00,0:09:11.94,Default,,0000,0000,0000,,political parties are trying to understand\Nhow we think, how we feel. So that's
Dialogue: 0,0:09:11.94,0:09:17.49,Default,,0000,0000,0000,,actually quite significant that you have\Nan app that's collecting information about
Dialogue: 0,0:09:17.49,0:09:24.11,Default,,0000,0000,0000,,how we feel on a daily basis. And\Nobviously, like when people enter all
Dialogue: 0,0:09:24.11,0:09:29.20,Default,,0000,0000,0000,,these data, their expectation at that\Npoint is that the data stays between
Dialogue: 0,0:09:29.20,0:09:35.48,Default,,0000,0000,0000,,between them and the app. And actually,\Nthere is very little in the privacy policy
Dialogue: 0,0:09:35.48,0:09:41.93,Default,,0000,0000,0000,,that could that would normally suggest\Nthat it was. So this is the moment where I
Dialogue: 0,0:09:41.93,0:09:45.71,Default,,0000,0000,0000,,actually should say we're not making this\Nup; like literally everything in this list
Dialogue: 0,0:09:45.71,0:09:51.75,Default,,0000,0000,0000,,of questions were things, literal terms,\Nthat they were asking. So we set out to
Dialogue: 0,0:09:51.75,0:09:55.40,Default,,0000,0000,0000,,look at the most popular menstruation\Napps. Do you want to carry on?
Dialogue: 0,0:09:55.40,0:09:59.84,Default,,0000,0000,0000,,Chris: Yeah. I forgot to introduce myself\Nas well. Really? That's a terrible
Dialogue: 0,0:09:59.84,0:10:02.44,Default,,0000,0000,0000,,speaking habit.\NEva: Christopher Weatherhead..
Dialogue: 0,0:10:02.44,0:10:08.74,Default,,0000,0000,0000,,Chris: .. Privacy International's\Ntechnology lead. So yeah.. What I said
Dialogue: 0,0:10:08.74,0:10:11.58,Default,,0000,0000,0000,,about our previous research, we have\Nactually looked at most of the very
Dialogue: 0,0:10:11.58,0:10:17.99,Default,,0000,0000,0000,,popular menstruation apps, the ones that\Nhave hundreds of thousands of downloads.
Dialogue: 0,0:10:17.99,0:10:21.91,Default,,0000,0000,0000,,And these apps - like as we're saying that\Nthis kind of work has been done before. A
Dialogue: 0,0:10:21.91,0:10:25.56,Default,,0000,0000,0000,,lot of these apps that come into quite a\Nlot of criticism, I'd spare you the free
Dialogue: 0,0:10:25.56,0:10:30.46,Default,,0000,0000,0000,,advertising about which ones particularly\Nbut most of them don't do anything
Dialogue: 0,0:10:30.46,0:10:36.50,Default,,0000,0000,0000,,particularly outrageous, at least between\Nthe app and the developers' servers. A lot
Dialogue: 0,0:10:36.50,0:10:39.47,Default,,0000,0000,0000,,of them don't share with third parties at\Nthat stage. So you can't look between the
Dialogue: 0,0:10:39.47,0:10:43.85,Default,,0000,0000,0000,,app and the server to see what they're\Nsharing. They might be sharing data from
Dialogue: 0,0:10:43.85,0:10:48.27,Default,,0000,0000,0000,,the developers' server to Facebook or to\Nother places but at least you can't see
Dialogue: 0,0:10:48.27,0:10:55.60,Default,,0000,0000,0000,,in-between. But we're an international\Norganization and we work around the globe.
Dialogue: 0,0:10:55.60,0:11:01.26,Default,,0000,0000,0000,,And most of the apps that get the most\Ndownloads are particularly Western, U.S.,
Dialogue: 0,0:11:01.26,0:11:07.70,Default,,0000,0000,0000,,European but they're not the most popular\Napps necessarily in a context like India
Dialogue: 0,0:11:07.70,0:11:12.81,Default,,0000,0000,0000,,and the Philippines and Latin America. So\Nwe thought we'd have a look and see those
Dialogue: 0,0:11:12.81,0:11:17.33,Default,,0000,0000,0000,,Apps. They're all available in Europe but\Nthey're not necessarily the most popular
Dialogue: 0,0:11:17.33,0:11:23.33,Default,,0000,0000,0000,,in Europe. And this is where things\Nstarted getting interesting. So what
Dialogue: 0,0:11:23.33,0:11:29.52,Default,,0000,0000,0000,,exactly did we do? Well, we started off by\Ntriaging through a large number of period
Dialogue: 0,0:11:29.52,0:11:36.27,Default,,0000,0000,0000,,trackers. And as Eva said earlier: every\Nlogo must be pink. And we were just kind
Dialogue: 0,0:11:36.27,0:11:40.42,Default,,0000,0000,0000,,of looking through to see how many\Ntrackers - this is using extras (?)
Dialogue: 0,0:11:40.42,0:11:46.60,Default,,0000,0000,0000,,privacy. We have our own instance in PI\Nand we just looked through to see how many
Dialogue: 0,0:11:46.60,0:11:50.78,Default,,0000,0000,0000,,trackers and who the trackers were. So,\Nfor example, this is Maya, which is
Dialogue: 0,0:11:50.78,0:11:54.52,Default,,0000,0000,0000,,exceptionally popular in India,\Npredominantly - it's made by an Indian
Dialogue: 0,0:11:54.52,0:12:01.05,Default,,0000,0000,0000,,company. And as you can see, it's got a\Nlarge number of trackers in it: a
Dialogue: 0,0:12:01.05,0:12:09.23,Default,,0000,0000,0000,,CleverTap, Facebook, Flurry, Google and\NInmobi? So we went through this process and
Dialogue: 0,0:12:09.23,0:12:14.78,Default,,0000,0000,0000,,this allowed us to cut down... There's\Nhundreds of period trackers. Not all of
Dialogue: 0,0:12:14.78,0:12:18.77,Default,,0000,0000,0000,,them are necessarily bad but it's nice to\Ntry to see which ones had the most
Dialogue: 0,0:12:18.77,0:12:24.50,Default,,0000,0000,0000,,trackers, where they were used and try and\Njust triage them a little bit. From this,
Dialogue: 0,0:12:24.50,0:12:33.19,Default,,0000,0000,0000,,we then run through PI's interception\Nenvironment, which is a VM that I've made.
Dialogue: 0,0:12:33.19,0:12:37.41,Default,,0000,0000,0000,,I actually made it last year for the talk\NI gave last year. And I said I'd release
Dialogue: 0,0:12:37.41,0:12:40.62,Default,,0000,0000,0000,,it after the talk and took me like three\Nmonths to release it but it's now
Dialogue: 0,0:12:40.62,0:12:45.42,Default,,0000,0000,0000,,available. You can go onto PI's website\Nand download it. It's a man in the middle
Dialogue: 0,0:12:45.42,0:12:52.86,Default,,0000,0000,0000,,proxy with a few settings - mainly for\Nlooking at iOS and Android apps to do data
Dialogue: 0,0:12:52.86,0:12:59.21,Default,,0000,0000,0000,,interception between them. And so we run\Nthrough that and we got to have a look at
Dialogue: 0,0:12:59.21,0:13:05.03,Default,,0000,0000,0000,,all the data that's being sent to and from\Nboth the app developer and third parties.
Dialogue: 0,0:13:05.03,0:13:10.81,Default,,0000,0000,0000,,And here's what we found.\NEva: So out of the six apps we looked out,
Dialogue: 0,0:13:10.81,0:13:17.92,Default,,0000,0000,0000,,five shared data with Facebook. Out of\Nthose five, three pinged Facebook to let
Dialogue: 0,0:13:17.92,0:13:23.99,Default,,0000,0000,0000,,them know when their users were\Ndownloading the app and opening the app.
Dialogue: 0,0:13:23.99,0:13:29.76,Default,,0000,0000,0000,,And that's already quite significant\Ninformation and we'll get to that later.
Dialogue: 0,0:13:29.76,0:13:37.06,Default,,0000,0000,0000,,Now, what's actually interesting and the\Nfocus of a report was on the two apps that
Dialogue: 0,0:13:37.06,0:13:42.04,Default,,0000,0000,0000,,shared every single piece of information\Nthat their users entered with Facebook and
Dialogue: 0,0:13:42.04,0:13:49.82,Default,,0000,0000,0000,,other third parties. So just to brief you:\Nthe two apps we focused on are both called
Dialogue: 0,0:13:49.82,0:13:55.33,Default,,0000,0000,0000,,Maya. So that's all very helpful. One is\Nspelled Maya: M-a-y-a. The other ones
Dialogue: 0,0:13:55.33,0:14:01.10,Default,,0000,0000,0000,,spellt Mia M-I-A. So, yeah, just bear with\Nme because this is actually quite
Dialogue: 0,0:14:01.10,0:14:09.80,Default,,0000,0000,0000,,confusing. But so initially we'll focus on\NMaya, which is - as Chris mentioned - an
Dialogue: 0,0:14:09.80,0:14:16.19,Default,,0000,0000,0000,,app that's based in India. There have a\Nuser base of several millions. Their are
Dialogue: 0,0:14:16.19,0:14:27.08,Default,,0000,0000,0000,,based in India. Userbase, mostly in India,\Nalso quite popular in the Philippines. So
Dialogue: 0,0:14:27.08,0:14:30.47,Default,,0000,0000,0000,,what's interesting with Maya is that they\Nstart sharing data with Facebook before
Dialogue: 0,0:14:30.47,0:14:34.80,Default,,0000,0000,0000,,you even get you agree to their privacy\Npolicy. So I should say already about the
Dialogue: 0,0:14:34.80,0:14:39.32,Default,,0000,0000,0000,,privacy policy of a lot of those apps that\Nwe looked at is that they are literally
Dialogue: 0,0:14:39.32,0:14:48.38,Default,,0000,0000,0000,,the definition of small prints. It's very\Nhard to read. It's legalese language. It
Dialogue: 0,0:14:48.38,0:14:53.62,Default,,0000,0000,0000,,really puts into perspective the whole\Nquestion of consent in GDPR because GDPR
Dialogue: 0,0:14:53.62,0:14:58.21,Default,,0000,0000,0000,,says like the consents must be informed.\NSo you must be able to understand what
Dialogue: 0,0:14:58.21,0:15:03.95,Default,,0000,0000,0000,,you're consenting to. When you're reading\Nthis extremely long, extremely opaque
Dialogue: 0,0:15:03.95,0:15:09.07,Default,,0000,0000,0000,,privacy policies of a lot - literally all\Nthe menstruation apps we've looked at,
Dialogue: 0,0:15:09.07,0:15:14.31,Default,,0000,0000,0000,,excluding one that didn't even bother\Nputting their privacy policy, actually.
Dialogue: 0,0:15:14.31,0:15:20.36,Default,,0000,0000,0000,,It's opaque. It's very hard to understand\Nand - absolutely, definitely, do not say
Dialogue: 0,0:15:20.36,0:15:25.48,Default,,0000,0000,0000,,that they're sharing information with\NFacebook. As I said, data sharing happened
Dialogue: 0,0:15:25.48,0:15:29.74,Default,,0000,0000,0000,,before you get to agree to their privacy\Npolicy. The other thing that's also worth
Dialogue: 0,0:15:29.74,0:15:33.49,Default,,0000,0000,0000,,remembering is that when to share\Ninformation with Facebook - doesn't matter
Dialogue: 0,0:15:33.49,0:15:39.18,Default,,0000,0000,0000,,if you have a Facebook account or not, the\Ninformation still being relayed. The other
Dialogue: 0,0:15:39.18,0:15:43.72,Default,,0000,0000,0000,,interesting thing that you'll notice as\Nwell in several of the slides is that the
Dialogue: 0,0:15:43.72,0:15:48.76,Default,,0000,0000,0000,,information that's being shared is tied to\Nyour identity through your unique ID
Dialogue: 0,0:15:48.76,0:15:54.64,Default,,0000,0000,0000,,identifiers, also your email address. But\Nbasically most of the questions we got
Dialogue: 0,0:15:54.64,0:16:00.22,Default,,0000,0000,0000,,when we released the research was like:\Noh, if I use a fake email address or if I
Dialogue: 0,0:16:00.22,0:16:06.08,Default,,0000,0000,0000,,use a fake name, is that OK? Well, it's\Nnot because even if you have a Facebook
Dialogue: 0,0:16:06.08,0:16:13.09,Default,,0000,0000,0000,,account through your unique identifier,\Nthey would definitely be able to trace you
Dialogue: 0,0:16:13.09,0:16:21.81,Default,,0000,0000,0000,,backs. There is no way to actually\Nanonymize this process unless - well at
Dialogue: 0,0:16:21.81,0:16:27.42,Default,,0000,0000,0000,,the end, unless you deliberately trying to\Ntrick it and use a separate phone
Dialogue: 0,0:16:27.42,0:16:34.04,Default,,0000,0000,0000,,basically for regular users. It's quite\Ndifficult. So this is what it looks like
Dialogue: 0,0:16:34.04,0:16:41.62,Default,,0000,0000,0000,,when you enter the data. So as I said, I\Ndidn't lie to you. This is the kind of
Dialogue: 0,0:16:41.62,0:16:49.34,Default,,0000,0000,0000,,questions they're asking you. And this is\Nwhat it looks like when it's being shared
Dialogue: 0,0:16:49.34,0:16:54.93,Default,,0000,0000,0000,,with Facebook. So you see the symptomes\Nchanging, for example, like blood
Dialogue: 0,0:16:54.93,0:17:00.34,Default,,0000,0000,0000,,pressure, swelling, acne, that's all being\Nshipped through craft out Facebook,
Dialogue: 0,0:17:00.34,0:17:06.35,Default,,0000,0000,0000,,through the Facebook SDK. This is what it\Nlooks like when they show you
Dialogue: 0,0:17:06.35,0:17:11.73,Default,,0000,0000,0000,,contraceptive practice, so again, like\Nwe're talking health data. Here we're
Dialogue: 0,0:17:11.73,0:17:17.89,Default,,0000,0000,0000,,talking sensitive data. We're talking\Nabout data that shouldn't normally require
Dialogue: 0,0:17:17.89,0:17:22.31,Default,,0000,0000,0000,,extra steps in terms of collecting it, in\Nterms of how it's being processed. But
Dialogue: 0,0:17:22.31,0:17:28.84,Default,,0000,0000,0000,,nope, in this case it was shared exactly\Nlike the rest. This's what it looks like.
Dialogue: 0,0:17:28.84,0:17:33.71,Default,,0000,0000,0000,,Well, so, yeah with sex life it was a\Nlittle bit different. So that's what it
Dialogue: 0,0:17:33.71,0:17:37.51,Default,,0000,0000,0000,,looks like when they're asking you about,\Nyou know, you just had sex, was it
Dialogue: 0,0:17:37.51,0:17:44.55,Default,,0000,0000,0000,,protected? Was it unprotected? The way it\Nwas shared with Facebook was a little bit
Dialogue: 0,0:17:44.55,0:17:51.49,Default,,0000,0000,0000,,cryptic, so to speak. So if you have\Nprotected sex, it was entered as love "2",
Dialogue: 0,0:17:51.49,0:17:57.78,Default,,0000,0000,0000,,unprotected sex was entered as Love "3". I\Nmanaged to figure that out pretty quickly.
Dialogue: 0,0:17:57.78,0:18:07.00,Default,,0000,0000,0000,,So it's not so cryptic. That's also quite\Nfunny. So Maya had a diary section where
Dialogue: 0,0:18:07.00,0:18:12.92,Default,,0000,0000,0000,,they encourage people to enter like their\Nnotes and your personal faults. And I
Dialogue: 0,0:18:12.92,0:18:18.68,Default,,0000,0000,0000,,mean, it's a menstruation app so you can\Nsort of get the idea of what people are
Dialogue: 0,0:18:18.68,0:18:21.90,Default,,0000,0000,0000,,going to be writing down in there or\Nexpected to write on. It's not going to be
Dialogue: 0,0:18:21.90,0:18:26.43,Default,,0000,0000,0000,,their shopping list, although shopping\Nlists could also be personal, sensitive,
Dialogue: 0,0:18:26.43,0:18:33.05,Default,,0000,0000,0000,,personal information, but.. So we were\Nwondering what would happen if you were to
Dialogue: 0,0:18:33.05,0:18:38.43,Default,,0000,0000,0000,,write in this in this diary and how this\Ndata would be processed. So we entered
Dialogue: 0,0:18:38.43,0:18:42.38,Default,,0000,0000,0000,,literally we entered something very\Nsensitive, entered here. This is what we
Dialogue: 0,0:18:42.38,0:18:53.41,Default,,0000,0000,0000,,wrote. And literally everything we wrote\Nwas shared with Facebook. Maya also shared
Dialogue: 0,0:18:53.41,0:18:58.08,Default,,0000,0000,0000,,your health data, not just with Facebook,\Nbut with a company called CleverTap that's
Dialogue: 0,0:18:58.08,0:19:05.44,Default,,0000,0000,0000,,based in California. So what's CleverTap?\NCleverTap is a data broker, basically.
Dialogue: 0,0:19:05.44,0:19:11.52,Default,,0000,0000,0000,,It's a company that - sort of similar to\NFacebook with the Facebook SDK. They
Dialogue: 0,0:19:11.52,0:19:16.95,Default,,0000,0000,0000,,expect of developers to hand over the data\Nand in exchange app developers get
Dialogue: 0,0:19:16.95,0:19:23.68,Default,,0000,0000,0000,,insights about like how people use the\Napp, what time of day. You know, the age
Dialogue: 0,0:19:23.68,0:19:30.79,Default,,0000,0000,0000,,of their users. They get all sorts of\Ninformation and analytics out of the data
Dialogue: 0,0:19:30.79,0:19:38.89,Default,,0000,0000,0000,,they share with this company. It took us\Nsome time to figure it out because it
Dialogue: 0,0:19:38.89,0:19:43.02,Default,,0000,0000,0000,,shared as wicked wizard?\NChris: Wicket Rocket.
Dialogue: 0,0:19:43.02,0:19:50.01,Default,,0000,0000,0000,,Eva: Wicket Rocket, yeah. But that's\Nexactly the same. Everything that was
Dialogue: 0,0:19:50.01,0:19:57.34,Default,,0000,0000,0000,,shared with Facebook was also shared with\NCleverTap again, with the email address
Dialogue: 0,0:19:57.34,0:20:04.99,Default,,0000,0000,0000,,that we were using - everything. Let's\Nshift. Now, let's look at the other Mia.
Dialogue: 0,0:20:04.99,0:20:10.11,Default,,0000,0000,0000,,It's not just the name that's similar,\Nit's also the data showing practices. Mia
Dialogue: 0,0:20:10.11,0:20:18.32,Default,,0000,0000,0000,,is based in Cypress, so in European Union.\NI should say, in all cases, regardless of
Dialogue: 0,0:20:18.32,0:20:22.12,Default,,0000,0000,0000,,where the company is based, the moment\Nthat they market the product in European
Dialogue: 0,0:20:22.12,0:20:29.46,Default,,0000,0000,0000,,Union, so like literally every app we\Nlooked at, they need to - well they should
Dialogue: 0,0:20:29.46,0:20:40.48,Default,,0000,0000,0000,,respect GDPR. Our European data protection\Nlaw. Now, the first thing that Mia asked
Dialogue: 0,0:20:40.48,0:20:44.94,Default,,0000,0000,0000,,when you started the app and again - I'll\Nget to that later about the significance
Dialogue: 0,0:20:44.94,0:20:49.71,Default,,0000,0000,0000,,of this - is why you're using the app or\Nyou using it to try and get pregnant or
Dialogue: 0,0:20:49.71,0:20:55.88,Default,,0000,0000,0000,,are you just using it to try to track your\Nperiods? Now, it's interesting because it
Dialogue: 0,0:20:55.88,0:21:00.07,Default,,0000,0000,0000,,doesn't change at all the way you interact\Nwith the app eventually. The apps stays
Dialogue: 0,0:21:00.07,0:21:05.18,Default,,0000,0000,0000,,exactly the same. But this is actually the\Nmost important kind of data. This is
Dialogue: 0,0:21:05.18,0:21:11.42,Default,,0000,0000,0000,,literally called the germ of data\Ncollection. It's trying to know when a
Dialogue: 0,0:21:11.42,0:21:15.97,Default,,0000,0000,0000,,woman is trying to get pregnant or not. So\Nthe reason this is the first question they
Dialogue: 0,0:21:15.97,0:21:21.39,Default,,0000,0000,0000,,ask is, well my guess on this is - they\Nwant to make sure that like even if you
Dialogue: 0,0:21:21.39,0:21:25.63,Default,,0000,0000,0000,,don't actually use the app that's at least\Nthat much information they can collect
Dialogue: 0,0:21:25.63,0:21:31.51,Default,,0000,0000,0000,,about you. And so this information was\Nshared immediately with Facebook and with
Dialogue: 0,0:21:31.51,0:21:36.53,Default,,0000,0000,0000,,AppsFlyer. AppsFlyer is very similar to\NCleverTap in the way it works. It's also a
Dialogue: 0,0:21:36.53,0:21:44.47,Default,,0000,0000,0000,,company that collects data from these apps\Nand that as services in terms of analytics
Dialogue: 0,0:21:44.47,0:21:54.48,Default,,0000,0000,0000,,and insights into user behavior. It's\Nbased in Israel. So this is what it looks
Dialogue: 0,0:21:54.48,0:22:04.71,Default,,0000,0000,0000,,like when you enter the information. Yeah,\Nmasturbation, pill. What kind of pill
Dialogue: 0,0:22:04.71,0:22:10.76,Default,,0000,0000,0000,,you're taking, your lifestyle habits. Now\Nwhere it's slightly different is that the
Dialogue: 0,0:22:10.76,0:22:15.96,Default,,0000,0000,0000,,information doesn't immediately get shared\Nwith Facebook but based on the information
Dialogue: 0,0:22:15.96,0:22:22.56,Default,,0000,0000,0000,,you enter, you get articles that are\Ntailored for you. So, for example, like
Dialogue: 0,0:22:22.56,0:22:27.36,Default,,0000,0000,0000,,when you select masturbation, you will\Nget, you know, masturbation: what you want
Dialogue: 0,0:22:27.36,0:22:35.85,Default,,0000,0000,0000,,to know but are ashamed to ask. Now,\Nwhat's eventually shared with Facebook is
Dialogue: 0,0:22:35.85,0:22:43.16,Default,,0000,0000,0000,,actually the kind of article that's being\Noffered to you. So basically, yes, the
Dialogue: 0,0:22:43.16,0:22:50.22,Default,,0000,0000,0000,,information is shared indirectly because\Nthen you know you have Facebook and...
Dialogue: 0,0:22:50.22,0:22:52.93,Default,,0000,0000,0000,,You've just entered masturbation because\Nyou're getting an article about
Dialogue: 0,0:22:52.93,0:22:58.94,Default,,0000,0000,0000,,masturbation. So this is what happened\Nwhen you enter alcohol. So expected
Dialogue: 0,0:22:58.94,0:23:02.63,Default,,0000,0000,0000,,effects of alcohol on a woman's body.\NThat's what happened when you enter
Dialogue: 0,0:23:02.63,0:23:06.15,Default,,0000,0000,0000,,"unprotected sex". So effectively, all the\Ninformation is still shared just
Dialogue: 0,0:23:06.15,0:23:14.44,Default,,0000,0000,0000,,indirectly through the articles you're\Ngetting. Yeah. Last thing also, I should
Dialogue: 0,0:23:14.44,0:23:18.45,Default,,0000,0000,0000,,say on this, in terms of the articles that\Nyou're getting, is that sometimes there
Dialogue: 0,0:23:18.45,0:23:23.49,Default,,0000,0000,0000,,was sort of also kind of like crossing the\Ndata.. was like.. so the articles will be
Dialogue: 0,0:23:23.49,0:23:30.48,Default,,0000,0000,0000,,about like: oh, you have cramps outside of\Nyour periods, for example, like during
Dialogue: 0,0:23:30.48,0:23:37.07,Default,,0000,0000,0000,,your fertile phase. And so you will get\Nthe article specifically for this and the
Dialogue: 0,0:23:37.07,0:23:42.56,Default,,0000,0000,0000,,information that's shared with Facebook\Nand with AppsFlyer is that this person is
Dialogue: 0,0:23:42.56,0:23:49.47,Default,,0000,0000,0000,,in their fertile period in this phase of\Ntheir cycles and having cramps. Now, why
Dialogue: 0,0:23:49.47,0:23:52.37,Default,,0000,0000,0000,,are menstruation apps so obsessed with\Nfinding out if you're trying to get
Dialogue: 0,0:23:52.37,0:23:59.84,Default,,0000,0000,0000,,pregnant? And so, this goes back to a lot\Nof the things I mentioned before that, you
Dialogue: 0,0:23:59.84,0:24:04.04,Default,,0000,0000,0000,,know, about wanting to know in the very\Nfirst place if you're trying to get
Dialogue: 0,0:24:04.04,0:24:10.26,Default,,0000,0000,0000,,pregnant or not. And also, this is\Nprobably why a lot of those apps are
Dialogue: 0,0:24:10.26,0:24:16.73,Default,,0000,0000,0000,,trying to really nail down in their\Nlanguage and discourse how you're using
Dialogue: 0,0:24:16.73,0:24:23.17,Default,,0000,0000,0000,,the apps for. When a person is pregnant,\Nthey're purchasing habit, their consumer
Dialogue: 0,0:24:23.17,0:24:29.91,Default,,0000,0000,0000,,habits change. Obviously, you know, you\Nbuy not only for yourself but you start
Dialogue: 0,0:24:29.91,0:24:36.67,Default,,0000,0000,0000,,buying for others as well. But also you're\Nbuying new things you've never purchased
Dialogue: 0,0:24:36.67,0:24:41.55,Default,,0000,0000,0000,,before. So what a regular person will be\Nquite difficult to change her purchasing
Dialogue: 0,0:24:41.55,0:24:47.55,Default,,0000,0000,0000,,habit was a person that's pregnant.\NThey'll be advertisers will be really keen
Dialogue: 0,0:24:47.55,0:24:52.87,Default,,0000,0000,0000,,to target them because this is a point of\Ntheir life where their habits change and
Dialogue: 0,0:24:52.87,0:24:58.44,Default,,0000,0000,0000,,where they can be more easily influenced\None way or another. So in other words,
Dialogue: 0,0:24:58.44,0:25:03.96,Default,,0000,0000,0000,,it's pink advertising time. In other more\Nwords and pictures, there's research done
Dialogue: 0,0:25:03.96,0:25:12.12,Default,,0000,0000,0000,,in 2014 in the US that was trying to sort\Nof evaluate the value of data for a
Dialogue: 0,0:25:12.12,0:25:19.32,Default,,0000,0000,0000,,person. So an average American person\Nthat's not pregnant was 10 cents. A person
Dialogue: 0,0:25:19.32,0:25:29.25,Default,,0000,0000,0000,,who's pregnant would be one dollar fifty.\NSo you may have noticed we using the past
Dialogue: 0,0:25:29.25,0:25:33.02,Default,,0000,0000,0000,,tense when we talked about - well I hope I\Ndid when I was speaking definitely into
Dialogue: 0,0:25:33.02,0:25:38.36,Default,,0000,0000,0000,,the lights at least - we used the past\Ntense when we talk about data sharing of
Dialogue: 0,0:25:38.36,0:25:43.33,Default,,0000,0000,0000,,these apps. That's because both Maya and\NMIA, which were the two apps we were
Dialogue: 0,0:25:43.33,0:25:47.98,Default,,0000,0000,0000,,really targeting with this report, stop\Nusing the Facebook SDK when we wrote to
Dialogue: 0,0:25:47.98,0:25:51.09,Default,,0000,0000,0000,,them about our research\Nbefore we published it.
Dialogue: 0,0:25:51.09,0:26:00.79,Default,,0000,0000,0000,,{\i1}applause{\i0}\NSo it was quite nice because he didn't
Dialogue: 0,0:26:00.79,0:26:05.69,Default,,0000,0000,0000,,even like rely on actually us publishing\Nthe report. It was merely at a stage of
Dialogue: 0,0:26:05.69,0:26:09.98,Default,,0000,0000,0000,,like, hey, this is all right of response.\NWe're gonna be publishing this. Do you
Dialogue: 0,0:26:09.98,0:26:13.55,Default,,0000,0000,0000,,have anything to say about this? And\Nessentially what they had to say is like:
Dialogue: 0,0:26:13.55,0:26:21.26,Default,,0000,0000,0000,,"Yep, sorry, apologies. We are stopping\Nthis." I think, you know.. What's really
Dialogue: 0,0:26:21.26,0:26:27.53,Default,,0000,0000,0000,,interesting as well to me about like the\Nhow quick the response was is.. it really
Dialogue: 0,0:26:27.53,0:26:34.16,Default,,0000,0000,0000,,shows how this is not a vital service for\Nthem. This is a plus. This is something
Dialogue: 0,0:26:34.16,0:26:41.68,Default,,0000,0000,0000,,that's a useful tool. But the fact that\Nthey immediately could just stop using it,
Dialogue: 0,0:26:41.68,0:26:48.27,Default,,0000,0000,0000,,I think really shows that, you know, it\Nwas.. I wouldn't see a lazy practice, but
Dialogue: 0,0:26:48.27,0:26:53.17,Default,,0000,0000,0000,,it's a case of light. As long as no one's\Ncomplaining, then you are going to carry
Dialogue: 0,0:26:53.17,0:27:00.30,Default,,0000,0000,0000,,on using it. And I think that was also the\Ndiscourse with your research. There was
Dialogue: 0,0:27:00.30,0:27:02.71,Default,,0000,0000,0000,,also a lot that changed\Ntheir behaviors after.
Dialogue: 0,0:27:02.71,0:27:06.50,Default,,0000,0000,0000,,Chris: A lot of the developers sometimes\Ndon't even realize necessarily what data
Dialogue: 0,0:27:06.50,0:27:12.01,Default,,0000,0000,0000,,they're up to sharing with people like\NFacebook, with people like CleverTap. They
Dialogue: 0,0:27:12.01,0:27:16.65,Default,,0000,0000,0000,,just integrate the SDK and\Nhope for the best.
Dialogue: 0,0:27:16.65,0:27:22.25,Default,,0000,0000,0000,,Eva: We also got this interesting response\Nfrom AppsFlyer is that it's very
Dialogue: 0,0:27:22.25,0:27:26.90,Default,,0000,0000,0000,,hypocritical. Essentially, what they're\Nsaying is like oh, like we specifically
Dialogue: 0,0:27:26.90,0:27:33.55,Default,,0000,0000,0000,,ask our customers or oh, yeah, do not\Nshare health data with us specifically for
Dialogue: 0,0:27:33.55,0:27:37.68,Default,,0000,0000,0000,,the reason I mentioned earlier, which is\Nwhat? Because of GDPR, you're normally
Dialogue: 0,0:27:37.68,0:27:44.52,Default,,0000,0000,0000,,expected to take extra step when you\Nprocess sensitive health data. So their
Dialogue: 0,0:27:44.52,0:27:48.81,Default,,0000,0000,0000,,response is that they as their customer to\Nnot share health data or sensitive
Dialogue: 0,0:27:48.81,0:27:54.90,Default,,0000,0000,0000,,personal data so they don't become liable\Nin terms of the law. So they were like,
Dialogue: 0,0:27:54.90,0:27:59.91,Default,,0000,0000,0000,,oh, we're sorry, like this is a breach of\Ncontract. Now, the reason is very
Dialogue: 0,0:27:59.91,0:28:04.29,Default,,0000,0000,0000,,hypocritical is that obviously when you\Nhave contracts with menstruation apps and
Dialogue: 0,0:28:04.29,0:28:07.86,Default,,0000,0000,0000,,actually Maya was not the only\Nmenstruation apps that we're working with.
Dialogue: 0,0:28:07.86,0:28:12.23,Default,,0000,0000,0000,,I mean, you know, what can you generally\Nexpect in terms of the kind of data you're
Dialogue: 0,0:28:12.23,0:28:19.14,Default,,0000,0000,0000,,gonna receive? So here's a conclusion for\Nus that research works. It's fun, it's
Dialogue: 0,0:28:19.14,0:28:26.98,Default,,0000,0000,0000,,easy to do. You know, Chris has not\Npublished the environment. It doesn't
Dialogue: 0,0:28:26.98,0:28:32.54,Default,,0000,0000,0000,,actually - once the environment is sort of\Nset up it doesn't actually require
Dialogue: 0,0:28:32.54,0:28:36.82,Default,,0000,0000,0000,,technical background, as you saw from the\Nslides it's pretty straightforward to
Dialogue: 0,0:28:36.82,0:28:41.96,Default,,0000,0000,0000,,actually understand how the data is being\Nshared. So you should do it, too. But more
Dialogue: 0,0:28:41.96,0:28:46.99,Default,,0000,0000,0000,,broadly, we think it's really important to\Ndo more research, not just at this stage
Dialogue: 0,0:28:46.99,0:28:54.27,Default,,0000,0000,0000,,of the process, but generally about the\Nsecurity and the data and the data showing
Dialogue: 0,0:28:54.27,0:29:00.14,Default,,0000,0000,0000,,practices of apps, because, you know, it's\Nhard law and more and more people are
Dialogue: 0,0:29:00.14,0:29:05.68,Default,,0000,0000,0000,,using or interacting with technology and\Nusing the Internet. So we need to do think
Dialogue: 0,0:29:05.68,0:29:10.51,Default,,0000,0000,0000,,much more carefully about the security\Nimplication of the apps we use and
Dialogue: 0,0:29:10.51,0:29:15.64,Default,,0000,0000,0000,,obviously it works. Thank you.
Dialogue: 0,0:29:15.64,0:29:25.37,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:29:25.37,0:29:29.52,Default,,0000,0000,0000,,Herald: Thank you. So, yeah, please line\Nup in front of the microphones. We can
Dialogue: 0,0:29:29.52,0:29:33.87,Default,,0000,0000,0000,,start with microphone two.\NMic 2: Hi. Thank you. So you mentioned
Dialogue: 0,0:29:33.87,0:29:39.12,Default,,0000,0000,0000,,that now we can check whether our data is\Nbeing shared with third parties on the
Dialogue: 0,0:29:39.12,0:29:42.46,Default,,0000,0000,0000,,path between the user and the developer.\NBut we cannot know for all the other apps
Dialogue: 0,0:29:42.46,0:29:46.28,Default,,0000,0000,0000,,and for these, what if it's not being\Nshared later from the developer, from the
Dialogue: 0,0:29:46.28,0:29:51.86,Default,,0000,0000,0000,,company to other companies. Have you\Nconceptualize some ways of testing that?
Dialogue: 0,0:29:51.86,0:29:55.66,Default,,0000,0000,0000,,Is it possible?\NChris: Yes. So you could do it, data
Dialogue: 0,0:29:55.66,0:30:03.98,Default,,0000,0000,0000,,separate access request and the GDPR that\Nwould... like the problem is it's quite
Dialogue: 0,0:30:03.98,0:30:11.30,Default,,0000,0000,0000,,hard to necessarily know. How the process\N- how the system outside of the app to
Dialogue: 0,0:30:11.30,0:30:16.14,Default,,0000,0000,0000,,serve relationship is quite hard to know\Nthe processes of that data and so it is
Dialogue: 0,0:30:16.14,0:30:20.31,Default,,0000,0000,0000,,quite opaque. They might apply a different\Nidentifier too, they might do other
Dialogue: 0,0:30:20.31,0:30:23.86,Default,,0000,0000,0000,,manipulations to that data so trying to\Ntrack down and prove this bit of data
Dialogue: 0,0:30:23.86,0:30:28.70,Default,,0000,0000,0000,,belong to you. It's quite challenging.\NEva: This is something we're going to try.
Dialogue: 0,0:30:28.70,0:30:32.07,Default,,0000,0000,0000,,We're going to be doing in 2020, actually.\NWe're going to be doing data subject
Dialogue: 0,0:30:32.07,0:30:38.33,Default,,0000,0000,0000,,access request of those apps that we've\Nbeen looking up to see if we find anything
Dialogue: 0,0:30:38.33,0:30:43.55,Default,,0000,0000,0000,,both under GDPR but also under different\Ndata protection laws in different
Dialogue: 0,0:30:43.55,0:30:49.98,Default,,0000,0000,0000,,countries. To see basically what we get,\Nhow much we can obtain from that.
Dialogue: 0,0:30:49.98,0:30:54.96,Default,,0000,0000,0000,,Herald: So I'd go with the signal angle.\NSignal: So what advice can you give us on
Dialogue: 0,0:30:54.96,0:31:00.33,Default,,0000,0000,0000,,how we can make people understand that\Nfrom a privacy perspective, it's better to
Dialogue: 0,0:31:00.33,0:31:05.28,Default,,0000,0000,0000,,use pen and paper instead of entering\Nsensitive data into any of these apps?
Dialogue: 0,0:31:05.28,0:31:10.44,Default,,0000,0000,0000,,Eva: I definitely wouldn't advise that. I\Nwouldn't advise pen and paper. I think for
Dialogue: 0,0:31:10.44,0:31:17.36,Default,,0000,0000,0000,,us like really the key... The work we are\Ndoing is not actually targeting users.
Dialogue: 0,0:31:17.36,0:31:21.28,Default,,0000,0000,0000,,It's targeting companies. We think it's\Ncompanies that really need to do better.
Dialogue: 0,0:31:21.28,0:31:26.27,Default,,0000,0000,0000,,We're often ask about, you know, advice to\Ncustomers or advice to users and
Dialogue: 0,0:31:26.27,0:31:32.03,Default,,0000,0000,0000,,consumers. But what I think and what we've\Nbeen telling companies as well is that,
Dialogue: 0,0:31:32.03,0:31:36.19,Default,,0000,0000,0000,,you know, their users trust you and they\Nhave the right to trust you. They also
Dialogue: 0,0:31:36.19,0:31:40.97,Default,,0000,0000,0000,,have the right to expect that you're\Nrespecting the law. The European Union has
Dialogue: 0,0:31:40.97,0:31:47.43,Default,,0000,0000,0000,,a very ambitious legislation when it comes\Nto privacy with GDPR. And so the least
Dialogue: 0,0:31:47.43,0:31:55.95,Default,,0000,0000,0000,,they can expect is that you're respecting\Nthe law. And so, no, I would ... and this
Dialogue: 0,0:31:55.95,0:31:59.54,Default,,0000,0000,0000,,is the thing, I think people have the\Nright to use those apps, they have the
Dialogue: 0,0:31:59.54,0:32:03.85,Default,,0000,0000,0000,,right to say, well, this is a useful\Nservice for me. It's really companies that
Dialogue: 0,0:32:03.85,0:32:08.21,Default,,0000,0000,0000,,need you. They need to up their game. They\Nneed to live up to the expectations of
Dialogue: 0,0:32:08.21,0:32:15.60,Default,,0000,0000,0000,,their consumers. Not the other way around.\NHerald: Microphone 1.
Dialogue: 0,0:32:15.60,0:32:19.22,Default,,0000,0000,0000,,Mic 1: Hi. So from the talk, it seems and\NI think that's what you get, you mostly
Dialogue: 0,0:32:19.22,0:32:23.32,Default,,0000,0000,0000,,focused on Android based apps. Can you\Nmaybe comment on what the situation is
Dialogue: 0,0:32:23.32,0:32:27.22,Default,,0000,0000,0000,,with iOS? Is there any technical\Ndifficulty or is it anything completely
Dialogue: 0,0:32:27.22,0:32:30.72,Default,,0000,0000,0000,,different with respect to these apps and\Napps in general?
Dialogue: 0,0:32:30.72,0:32:33.67,Default,,0000,0000,0000,,Chris: There's not really a technical\Ndifficulty like the setup a little bit
Dialogue: 0,0:32:33.67,0:32:38.80,Default,,0000,0000,0000,,different, but functionally you can look\Nat the same kind of data. The focus here,
Dialogue: 0,0:32:38.80,0:32:44.96,Default,,0000,0000,0000,,though, is also.. So it's two-fold in some\Nrespects. Most of the places that these
Dialogue: 0,0:32:44.96,0:32:49.94,Default,,0000,0000,0000,,apps are used are heavily dominated\NAndroid territories, places like India,
Dialogue: 0,0:32:49.94,0:32:55.53,Default,,0000,0000,0000,,the Philippines. iOS penetration there,\Nuh, Apple device penetration there is very
Dialogue: 0,0:32:55.53,0:33:01.98,Default,,0000,0000,0000,,low. There's no technical reason not to\Nlook at Apple devices. But like in this
Dialogue: 0,0:33:01.98,0:33:06.78,Default,,0000,0000,0000,,particular context, it's not necessarily\Nhugely relevant. So does that answer your
Dialogue: 0,0:33:06.78,0:33:08.99,Default,,0000,0000,0000,,question?\NMic 1: And technically with youre set-up,
Dialogue: 0,0:33:08.99,0:33:12.06,Default,,0000,0000,0000,,you could also do the same\Nanalysis with an iOS device?
Dialogue: 0,0:33:12.06,0:33:17.34,Default,,0000,0000,0000,,Chris: Yeah. As I said it's a little bit\Nof a change to how you... You have to
Dialogue: 0,0:33:17.34,0:33:22.49,Default,,0000,0000,0000,,register the device as an MDM dev.. like a\Nmobile profile device. Otherwise you can
Dialogue: 0,0:33:22.49,0:33:30.81,Default,,0000,0000,0000,,do the exact same level of interception.\NMic: Uh, hi. My question is actually
Dialogue: 0,0:33:30.81,0:33:33.21,Default,,0000,0000,0000,,related to the last question\Nis a little bit technical.
Dialogue: 0,0:33:33.21,0:33:35.62,Default,,0000,0000,0000,,Chris: Sure.\NMic: I'm also doing some research on apps
Dialogue: 0,0:33:35.62,0:33:39.54,Default,,0000,0000,0000,,and I've noticed with the newest versions\Nof Android that they're making more
Dialogue: 0,0:33:39.54,0:33:44.29,Default,,0000,0000,0000,,difficult to install custom certificates\Nto have this pass- through and check what
Dialogue: 0,0:33:44.29,0:33:49.07,Default,,0000,0000,0000,,the apps are actually communicating to\Ntheir home servers. Have you find a way to
Dialogue: 0,0:33:49.07,0:33:54.03,Default,,0000,0000,0000,,make this easier?\NChris: Yes. So we actually hit the same
Dialogue: 0,0:33:54.03,0:34:01.54,Default,,0000,0000,0000,,issue as you in some respects. So the\Ninstalling of custom certificates was not
Dialogue: 0,0:34:01.54,0:34:05.55,Default,,0000,0000,0000,,really an obstacle because you can add to\Nthe user if it's a rich device, you can
Dialogue: 0,0:34:05.55,0:34:13.51,Default,,0000,0000,0000,,add them to the system store and they are\Ntrusted by all the apps on the device. The
Dialogue: 0,0:34:13.51,0:34:19.33,Default,,0000,0000,0000,,problem we're now hitting is the Android 9\Nand 10 have TLS 1.3 and TLS 1.3
Dialogue: 0,0:34:19.33,0:34:24.34,Default,,0000,0000,0000,,to text as a man in the middle or at\Nleast it tries to might terminate the
Dialogue: 0,0:34:24.34,0:34:28.76,Default,,0000,0000,0000,,connection. Uh, this is a bit of a\Nproblem. So currently all our research is
Dialogue: 0,0:34:28.76,0:34:37.49,Default,,0000,0000,0000,,still running on Android 8.1 devices. This\Nisn't going to be sustainable long term.
Dialogue: 0,0:34:37.49,0:34:43.21,Default,,0000,0000,0000,,Herald: Um, 4.\NMic 4: Hey, thank you for the great talk.
Dialogue: 0,0:34:43.21,0:34:47.25,Default,,0000,0000,0000,,Your research is obviously targeted in a\Nconstructive, critical way towards
Dialogue: 0,0:34:47.25,0:34:53.25,Default,,0000,0000,0000,,companies that are making apps surrounding\Nmenstrual research. Did you learn anything
Dialogue: 0,0:34:53.25,0:34:57.21,Default,,0000,0000,0000,,from this context that you would want to\Npass on to people who research this area
Dialogue: 0,0:34:57.21,0:35:03.36,Default,,0000,0000,0000,,more generally? I'm thinking, for example,\Nof Paramount Corp in the US, who've done
Dialogue: 0,0:35:03.36,0:35:07.70,Default,,0000,0000,0000,,micro dosing research on LSD and are\Nstarting a breakout study on menstrual
Dialogue: 0,0:35:07.70,0:35:12.08,Default,,0000,0000,0000,,issues.\NEva: Well, I think this is why I was
Dialogue: 0,0:35:12.08,0:35:15.98,Default,,0000,0000,0000,,concluded on it. I think there is a\Nthere's still a lot of research that needs
Dialogue: 0,0:35:15.98,0:35:21.09,Default,,0000,0000,0000,,to be done in terms of the sharing. And\Nobviously, I think anything that touches
Dialogue: 0,0:35:21.09,0:35:27.83,Default,,0000,0000,0000,,on people's health is a key priority\Nbecause it's something people relate very
Dialogue: 0,0:35:27.83,0:35:33.75,Default,,0000,0000,0000,,strongly to. The consequences, especially\Nin the US, for example, of sharing health
Dialogue: 0,0:35:33.75,0:35:38.70,Default,,0000,0000,0000,,data like this, of having - you know -\Ndata, even like your blood pressure and so
Dialogue: 0,0:35:38.70,0:35:42.76,Default,,0000,0000,0000,,on. Like what are the consequences if\Nthose informations are gonna be shared,
Dialogue: 0,0:35:42.76,0:35:46.59,Default,,0000,0000,0000,,for example, with like insurance companies\Nand so on. This is what I think is
Dialogue: 0,0:35:46.59,0:35:52.47,Default,,0000,0000,0000,,absolutely essential to have a better\Nunderstanding of the data collection and
Dialogue: 0,0:35:52.47,0:35:57.57,Default,,0000,0000,0000,,sharing practices of the services. The\Nmoments when you have health data that's
Dialogue: 0,0:35:57.57,0:35:59.72,Default,,0000,0000,0000,,being involved.\NChris: .. yeah because we often focus
Dialogue: 0,0:35:59.72,0:36:06.00,Default,,0000,0000,0000,,about this being an advertising issue. But\Nin that sense as well, insurance and even
Dialogue: 0,0:36:06.00,0:36:09.95,Default,,0000,0000,0000,,credit referencing of all sorts of other\Nthings become problematic, especially when
Dialogue: 0,0:36:09.95,0:36:14.75,Default,,0000,0000,0000,,it comes to pregnancy related.\NEva: Yeah, even employers could be after
Dialogue: 0,0:36:14.75,0:36:18.51,Default,,0000,0000,0000,,this kind of information.\NHerald: Six.
Dialogue: 0,0:36:18.51,0:36:24.45,Default,,0000,0000,0000,,Mic 6: Hi. I'm wondering if there is an\Neasy way or a tool which we can use to
Dialogue: 0,0:36:24.45,0:36:32.58,Default,,0000,0000,0000,,detect if apps are using our data or are\Nreporting them to Facebook or whatever. Or
Dialogue: 0,0:36:32.58,0:36:39.83,Default,,0000,0000,0000,,if we can even use those apps but block\Nthis data from being reported to Facebook.
Dialogue: 0,0:36:39.83,0:36:45.65,Default,,0000,0000,0000,,Chris: Yes. So, you can file all of faith\Ngraft on Facebook.com and stop sending
Dialogue: 0,0:36:45.65,0:36:51.77,Default,,0000,0000,0000,,data to that. There's a few issues here.\NFirstly, it doesn't really like.. This
Dialogue: 0,0:36:51.77,0:36:57.94,Default,,0000,0000,0000,,audience can do this. Most users don't\Nhave the technical nuance to know what
Dialogue: 0,0:36:57.94,0:37:02.39,Default,,0000,0000,0000,,needs to be blocked, what doesn't\Nnecessarily need to be blocked. It's on
Dialogue: 0,0:37:02.39,0:37:07.30,Default,,0000,0000,0000,,the companies to be careful with users\Ndata. It's not up to the users to try and
Dialogue: 0,0:37:07.30,0:37:13.50,Default,,0000,0000,0000,,defend against.. It shouldn't be on the\Nuse to defend against malicious data
Dialogue: 0,0:37:13.50,0:37:17.49,Default,,0000,0000,0000,,sharing or...\NEva: You know... also one interesting
Dialogue: 0,0:37:17.49,0:37:21.93,Default,,0000,0000,0000,,thing is that if Facebook had put this in\Nplace of light where you could opt out
Dialogue: 0,0:37:21.93,0:37:25.47,Default,,0000,0000,0000,,from data sharing with the apps you're\Nusing, but that only works if you're a
Dialogue: 0,0:37:25.47,0:37:29.84,Default,,0000,0000,0000,,Facebook user. And as I said, like this\Ndata has been collected whether you are a
Dialogue: 0,0:37:29.84,0:37:34.23,Default,,0000,0000,0000,,user or not. So in a sense, for people who\Naren't Facebook users, they couldn't opt
Dialogue: 0,0:37:34.23,0:37:37.72,Default,,0000,0000,0000,,out of this.\NChris: The Facebook SDK the developers are
Dialogue: 0,0:37:37.72,0:37:46.69,Default,,0000,0000,0000,,integrating the default state for sharing\Nof data is on, the flag is true. And
Dialogue: 0,0:37:46.69,0:37:56.48,Default,,0000,0000,0000,,although they have a long legal text on\Nthe help pages for the developer tools,
Dialogue: 0,0:37:56.48,0:38:00.54,Default,,0000,0000,0000,,it's like unless you have a decent\Nunderstanding of local data protection
Dialogue: 0,0:38:00.54,0:38:04.89,Default,,0000,0000,0000,,practice or local protection law. It's\Nlike it's not it's not something that most
Dialogue: 0,0:38:04.89,0:38:08.84,Default,,0000,0000,0000,,developers are gonna be able to understand\Nwhy this flag should be something
Dialogue: 0,0:38:08.84,0:38:16.32,Default,,0000,0000,0000,,different from on. You know there's loads\Nof flags in the SDK, which flags should be
Dialogue: 0,0:38:16.32,0:38:21.93,Default,,0000,0000,0000,,on and off, depending on which\Njurisdiction you're selling to, or users
Dialogue: 0,0:38:21.93,0:38:27.24,Default,,0000,0000,0000,,going to be in.\NHerald: Signal Angel, again.
Dialogue: 0,0:38:27.24,0:38:31.53,Default,,0000,0000,0000,,Singal: Do you know any good apps which\Ndon't share data and are privacy friendly?
Dialogue: 0,0:38:31.53,0:38:37.12,Default,,0000,0000,0000,,Probably even one that is open source.\NEva: So, I mean, as in the problem which
Dialogue: 0,0:38:37.12,0:38:43.26,Default,,0000,0000,0000,,is why I wouldn't want to vouch for any\Napp is that even in the apps that, you
Dialogue: 0,0:38:43.26,0:38:48.50,Default,,0000,0000,0000,,know, where in terms of like the traffic\Nanalysis we've done, we didn't see any any
Dialogue: 0,0:38:48.50,0:38:53.16,Default,,0000,0000,0000,,data sharing. As Chris was explaining, the\Ndata can be shared at a later stage and
Dialogue: 0,0:38:53.16,0:39:00.72,Default,,0000,0000,0000,,it'd be impossible for us to really find\Nout. So.. no, I can't be vouching for any
Dialogue: 0,0:39:00.72,0:39:04.65,Default,,0000,0000,0000,,app. I don't know if you can...\NChris: The problem is we can't ever look
Dialogue: 0,0:39:04.65,0:39:10.81,Default,,0000,0000,0000,,like one specific moment in time to see\Nwhether data is being shared, unlike what
Dialogue: 0,0:39:10.81,0:39:17.69,Default,,0000,0000,0000,,was good today might bad tomorrow. What\Nwas bad yesterday might be good today.
Dialogue: 0,0:39:17.69,0:39:25.23,Default,,0000,0000,0000,,Although, I was in Argentina recently\Nspeaking to a group of feminist activists,
Dialogue: 0,0:39:25.23,0:39:31.86,Default,,0000,0000,0000,,and they have been developing a\Nmenstruation tracking app. And the app was
Dialogue: 0,0:39:31.86,0:39:37.80,Default,,0000,0000,0000,,removed from the Google Play store because\Nit had illustrations that were deemed
Dialogue: 0,0:39:37.80,0:39:42.50,Default,,0000,0000,0000,,pornographic. But they were illustrations\Naround medical related stuff. So even
Dialogue: 0,0:39:42.50,0:39:45.17,Default,,0000,0000,0000,,people, who were trying to do the right\Nthing, going through the open source
Dialogue: 0,0:39:45.17,0:39:49.72,Default,,0000,0000,0000,,channels are still fighting a completely\Ndifferent issue when it comes to
Dialogue: 0,0:39:49.72,0:39:52.94,Default,,0000,0000,0000,,menstruation tracking.\NIt's a very fine line.
Dialogue: 0,0:39:52.94,0:39:57.33,Default,,0000,0000,0000,,Herald: Um, three.\N{\i1}inaudible{\i0}
Dialogue: 0,0:39:57.33,0:40:01.77,Default,,0000,0000,0000,,Eva: Sorry, can't hear -the Mic's not\Nworking.
Dialogue: 0,0:40:01.77,0:40:04.79,Default,,0000,0000,0000,,Herald: Microphone three.\NMic 3: Test.
Dialogue: 0,0:40:04.79,0:40:09.85,Default,,0000,0000,0000,,Eva: Yeah, it's great - perfect.\NMic 3: I was wondering if the graph API
Dialogue: 0,0:40:09.85,0:40:16.56,Default,,0000,0000,0000,,endpoint was actually in place to trick\Nmenstruation data or is it more like a
Dialogue: 0,0:40:16.56,0:40:22.97,Default,,0000,0000,0000,,general purpose advertisement\Ntracking thing or. Yeah.
Dialogue: 0,0:40:22.97,0:40:29.36,Default,,0000,0000,0000,,Chris: So my understanding is that there's\Ntwo broad kinds of data that Facebook gets
Dialogue: 0,0:40:29.36,0:40:35.97,Default,,0000,0000,0000,,as automated app events that Facebook were\Naware of. So app open, app close, app
Dialogue: 0,0:40:35.97,0:40:41.76,Default,,0000,0000,0000,,install, relinking. Relinking is quite an\Nimportant one for Facebook. That way they
Dialogue: 0,0:40:41.76,0:40:44.94,Default,,0000,0000,0000,,check to see whether you already have a\NFacebook account logged in to log the app
Dialogue: 0,0:40:44.94,0:40:49.95,Default,,0000,0000,0000,,to your Facebook account when standing.\NThere's also a load of custom events that
Dialogue: 0,0:40:49.95,0:40:55.40,Default,,0000,0000,0000,,the app developers can put in. There is\Nthen collated back to a data set - I would
Dialogue: 0,0:40:55.40,0:41:01.52,Default,,0000,0000,0000,,imagine on the other side. So when it\Ncomes to things like whether it's nausea
Dialogue: 0,0:41:01.52,0:41:06.39,Default,,0000,0000,0000,,or some of the other health issues, it is\Nactually being cross-referenced by the
Dialogue: 0,0:41:06.39,0:41:11.82,Default,,0000,0000,0000,,developer. Does that answer your question?\NMic 3: Yes, thank you.
Dialogue: 0,0:41:11.82,0:41:16.32,Default,,0000,0000,0000,,Herald: Five, microphone five.\NMic 5: Can you repeat what you said in the
Dialogue: 0,0:41:16.32,0:41:23.29,Default,,0000,0000,0000,,beginning about the menstruation apps used\Nin Europe, especially Clue and the Period
Dialogue: 0,0:41:23.29,0:41:29.86,Default,,0000,0000,0000,,Tracker? Yeah. So those are the most\Npopular apps actually across the world,
Dialogue: 0,0:41:29.86,0:41:35.10,Default,,0000,0000,0000,,not just in Europe and the US. A lot of\Nthem in terms of like the traffic analysis
Dialogue: 0,0:41:35.10,0:41:40.98,Default,,0000,0000,0000,,stage, a lot of them have not clean up\Ntheir app. So we can't see any any data
Dialogue: 0,0:41:40.98,0:41:46.09,Default,,0000,0000,0000,,sharing happening at that stage. But as I\Nsaid, I can't be vouching for them and
Dialogue: 0,0:41:46.09,0:41:49.68,Default,,0000,0000,0000,,saying, oh, yeah, those are safe and fine\Nto use because we don't know what's
Dialogue: 0,0:41:49.68,0:41:54.31,Default,,0000,0000,0000,,actually happening to the data once it's\Nbeen collected by the app. All we can say
Dialogue: 0,0:41:54.31,0:42:01.87,Default,,0000,0000,0000,,is that as far as the research we've done\Ngoes, we didn't see any data being shed
Dialogue: 0,0:42:01.87,0:42:06.75,Default,,0000,0000,0000,,Chris: Those apps you mentioned have been\Ninvestigated by The Wall Street Journal
Dialogue: 0,0:42:06.75,0:42:11.79,Default,,0000,0000,0000,,and The New York Times relatively\Nrecently. So they've been.. had quite like
Dialogue: 0,0:42:11.79,0:42:15.72,Default,,0000,0000,0000,,a spotlight on them. So they've had to\Nreally up their game and a lot of ways
Dialogue: 0,0:42:15.72,0:42:20.59,Default,,0000,0000,0000,,which we would like everyone to do. But as\NEva says, we don't know what else they
Dialogue: 0,0:42:20.59,0:42:24.74,Default,,0000,0000,0000,,might be doing with that data on their\Nside, not necessarily between the phone
Dialogue: 0,0:42:24.74,0:42:29.15,Default,,0000,0000,0000,,and the server but from their server to\Nanother server.
Dialogue: 0,0:42:29.15,0:42:32.51,Default,,0000,0000,0000,,Herald: Microphone one.\NMic 1: Hi. Thank you for the insightful
Dialogue: 0,0:42:32.51,0:42:37.62,Default,,0000,0000,0000,,talk. I have a question that goes in a\Nsimilar direction. Do you know whether or
Dialogue: 0,0:42:37.62,0:42:44.08,Default,,0000,0000,0000,,not these apps, even if they adhere to\NGDPR rules collect the data to then at a
Dialogue: 0,0:42:44.08,0:42:48.85,Default,,0000,0000,0000,,later point at least sell it to the\Nhighest bidder? Because a lot of them are
Dialogue: 0,0:42:48.85,0:42:53.16,Default,,0000,0000,0000,,free to use. And I wonder what is their\Nmain goal besides that?
Dialogue: 0,0:42:53.16,0:42:58.44,Default,,0000,0000,0000,,Eva: I mean, the advertisement his how\Nthey make profit. And so, I mean, the
Dialogue: 0,0:42:58.44,0:43:04.45,Default,,0000,0000,0000,,whole question about them trying to know\Nif you're pregnant or not is that this
Dialogue: 0,0:43:04.45,0:43:11.54,Default,,0000,0000,0000,,information can eventually be - you know -\Nbe monetized through, you know, through
Dialogue: 0,0:43:11.54,0:43:17.07,Default,,0000,0000,0000,,how they target the advertisement at you.\NActually when you're using those apps, you
Dialogue: 0,0:43:17.07,0:43:20.34,Default,,0000,0000,0000,,can see in some of the slides, like you're\Nconstantly like being flushed with like
Dialogue: 0,0:43:20.34,0:43:25.63,Default,,0000,0000,0000,,all sorts of advertisement on the app, you\Nknow, whether they are selling it
Dialogue: 0,0:43:25.63,0:43:31.47,Default,,0000,0000,0000,,externally or not - I can't tell. But what\NI can tell is, yeah, your business model
Dialogue: 0,0:43:31.47,0:43:34.96,Default,,0000,0000,0000,,is advertisement and so they are deriving\Nprofit from the data they collect.
Dialogue: 0,0:43:34.96,0:43:40.41,Default,,0000,0000,0000,,Absolutely.\NHerald: Again, on microphone one.
Dialogue: 0,0:43:40.41,0:43:44.60,Default,,0000,0000,0000,,Mic 1: Thank you. I was wondering if there\Nwas more of a big data kind of aspect to
Dialogue: 0,0:43:44.60,0:43:50.08,Default,,0000,0000,0000,,it as well, because these are really\Ninteresting medical information on women’s
Dialogue: 0,0:43:50.08,0:43:54.56,Default,,0000,0000,0000,,cycles in general.\NEva: Yeah, and the answer is, like, I call
Dialogue: 0,0:43:54.56,0:43:58.03,Default,,0000,0000,0000,,it—this is a bit of a black box and\Nespecially in the way, for example, that
Dialogue: 0,0:43:58.03,0:44:03.10,Default,,0000,0000,0000,,Facebook is using this data like we don't\Nknow. We can assume that this is like part
Dialogue: 0,0:44:03.10,0:44:07.28,Default,,0000,0000,0000,,of the … we could assume this is part of\Nthe profiling that Facebook does of both
Dialogue: 0,0:44:07.28,0:44:13.40,Default,,0000,0000,0000,,their users and their non-users. But the\Nway the way this data is actually
Dialogue: 0,0:44:13.40,0:44:19.51,Default,,0000,0000,0000,,processed also by those apps through data\Nbrokers and so on, it’s a bit of a black
Dialogue: 0,0:44:19.51,0:44:27.53,Default,,0000,0000,0000,,box.\NHerald: Microphone 1.
Dialogue: 0,0:44:27.53,0:44:32.03,Default,,0000,0000,0000,,Mic 1: Yeah. Thank you a lot for your talk\Nand I have two completely different
Dialogue: 0,0:44:32.03,0:44:37.63,Default,,0000,0000,0000,,questions. The first one is: you've been\Nfocusing a lot on advertising and how this
Dialogue: 0,0:44:37.63,0:44:44.94,Default,,0000,0000,0000,,data is used to sell to advertisers. But I\Nmean, like you aim to be pregnant or not.
Dialogue: 0,0:44:44.94,0:44:48.81,Default,,0000,0000,0000,,It's like it has to be the best kept\Nsecret, at least in Switzerland for any
Dialogue: 0,0:44:48.81,0:44:54.43,Default,,0000,0000,0000,,female person, because like if you also\Nwant to get employed, your employer must
Dialogue: 0,0:44:54.43,0:44:59.74,Default,,0000,0000,0000,,not know whether or not you want to get\Npregnant. And so I would like to ask,
Dialogue: 0,0:44:59.74,0:45:06.23,Default,,0000,0000,0000,,like, how likely is it that this kind of\Ndata is also potentially sold to employers
Dialogue: 0,0:45:06.23,0:45:12.00,Default,,0000,0000,0000,,who may want to poke into your health and\Nreproductive situation? And then my other
Dialogue: 0,0:45:12.00,0:45:17.29,Default,,0000,0000,0000,,question is entirely different, because we\Nalso know that female health is one of the
Dialogue: 0,0:45:17.29,0:45:22.22,Default,,0000,0000,0000,,least researched topics around, and that's\Nactually a huge problem. Like so little is
Dialogue: 0,0:45:22.22,0:45:27.51,Default,,0000,0000,0000,,actually known about female health and the\Nkind of data that these apps collect is
Dialogue: 0,0:45:27.51,0:45:34.31,Default,,0000,0000,0000,,actually a gold mine to advance research\Non health issues that are specific for
Dialogue: 0,0:45:34.31,0:45:38.92,Default,,0000,0000,0000,,certain bodies like female bodies. And so\NI would also like to know like how would
Dialogue: 0,0:45:38.92,0:45:43.86,Default,,0000,0000,0000,,it be possible to still gather this kind\Nof data and still to collect it, but use
Dialogue: 0,0:45:43.86,0:45:48.49,Default,,0000,0000,0000,,it for like a beneficial purpose, like it\Nto improve knowledge on these issues?
Dialogue: 0,0:45:48.49,0:45:53.69,Default,,0000,0000,0000,,Eva: Sure. So to answer your first\Nquestion, the answer will be similar to
Dialogue: 0,0:45:53.69,0:45:58.30,Default,,0000,0000,0000,,the previous answer I gave, which is, you\Nknow, it's black box problem. It's like
Dialogue: 0,0:45:58.30,0:46:02.41,Default,,0000,0000,0000,,it's very difficult to know exactly, you\Nknow, what's actually happening to this
Dialogue: 0,0:46:02.41,0:46:08.57,Default,,0000,0000,0000,,data. Obviously, GDPR is there to prevent\Nsomething from happening. But as we've
Dialogue: 0,0:46:08.57,0:46:17.89,Default,,0000,0000,0000,,seen from these apps, like they were, you\Nknow, towing a very blurry line. And so
Dialogue: 0,0:46:17.89,0:46:22.36,Default,,0000,0000,0000,,the risk, obviously, of … this is\Nsomething that can’t be relia…. I can't be
Dialogue: 0,0:46:22.36,0:46:26.29,Default,,0000,0000,0000,,saying, oh, this is happening because I\Nhave no evidence that this is happening.
Dialogue: 0,0:46:26.29,0:46:31.76,Default,,0000,0000,0000,,But obviously, the risk of multiple, the\Nrisk of like employers, as you say, the
Dialogue: 0,0:46:31.76,0:46:36.49,Default,,0000,0000,0000,,insurance companies that could get it,\Nthat political parties could get it and
Dialogue: 0,0:46:36.49,0:46:40.96,Default,,0000,0000,0000,,target their messages based on information\Nthey have about your mood, about, you
Dialogue: 0,0:46:40.96,0:46:45.26,Default,,0000,0000,0000,,know, even the fact that you're trying to\Nstart a family. So, yeah, there is a very
Dialogue: 0,0:46:45.26,0:46:50.24,Default,,0000,0000,0000,,broad range of risk. The advertisement we\Nknow for sure is happening because this is
Dialogue: 0,0:46:50.24,0:46:55.85,Default,,0000,0000,0000,,like the basis of their business model.\NThe risk, the range of risk is very, very
Dialogue: 0,0:46:55.85,0:46:59.94,Default,,0000,0000,0000,,broad.\NChris: To just expand on that: Again, as
Dialogue: 0,0:46:59.94,0:47:05.43,Default,,0000,0000,0000,,Eva said, we can't point out a specific\Nexample of any of this. But if you look at
Dialogue: 0,0:47:05.43,0:47:10.26,Default,,0000,0000,0000,,some of the other data brokers, her\Nexperience as a data broker, they collect.
Dialogue: 0,0:47:10.26,0:47:16.35,Default,,0000,0000,0000,,They have a statutory response. In the UK\Nis a statutory job of being a credit
Dialogue: 0,0:47:16.35,0:47:23.52,Default,,0000,0000,0000,,reference agency, but they also run what\Nis believed to be armed data enrichment.
Dialogue: 0,0:47:23.52,0:47:29.20,Default,,0000,0000,0000,,One of the things her employers could do\Nis by experience data to when hiring
Dialogue: 0,0:47:29.20,0:47:35.69,Default,,0000,0000,0000,,staff. Like I can't say that if this data\Never ends up there. But, you know, as they
Dialogue: 0,0:47:35.69,0:47:41.12,Default,,0000,0000,0000,,all collect, there is people collecting\Ndata and using it for some level of
Dialogue: 0,0:47:41.12,0:47:45.45,Default,,0000,0000,0000,,auditing.\NEva: And to transfer your second question.
Dialogue: 0,0:47:45.45,0:47:49.81,Default,,0000,0000,0000,,I think this is a very important problem\Nyou point out is the question of data
Dialogue: 0,0:47:49.81,0:47:56.23,Default,,0000,0000,0000,,inequality and whose data gets collected\Nfor what purpose. There is I do quite a
Dialogue: 0,0:47:56.23,0:48:01.10,Default,,0000,0000,0000,,lot of work on delivery of state services.\NFor example, when there are populations
Dialogue: 0,0:48:01.10,0:48:05.94,Default,,0000,0000,0000,,that are isolated, not using technology\Nand so on. You might just be missing out
Dialogue: 0,0:48:05.94,0:48:12.45,Default,,0000,0000,0000,,on people, for example, who should be in\Nneed of health care or state
Dialogue: 0,0:48:12.45,0:48:18.12,Default,,0000,0000,0000,,support and so on. Just because you like\Ndata about about them. And so, female
Dialogue: 0,0:48:18.12,0:48:24.26,Default,,0000,0000,0000,,health is obviously a very key issue. We\Njust, we literally lack sufficient health
Dialogue: 0,0:48:24.26,0:48:30.52,Default,,0000,0000,0000,,data about about woman on women's health\Nspecifically. Now, in terms of how data is
Dialogue: 0,0:48:30.52,0:48:35.55,Default,,0000,0000,0000,,processed in medical research, then\Nthere's actually protocol a in place
Dialogue: 0,0:48:35.55,0:48:40.47,Default,,0000,0000,0000,,normally to ensure, to ensure consent, to\Nensure explicit consent, to ensure that
Dialogue: 0,0:48:40.47,0:48:47.21,Default,,0000,0000,0000,,the data is properly collected. And so I\Nthink I wouldn't want you means that you,
Dialogue: 0,0:48:47.21,0:48:52.01,Default,,0000,0000,0000,,just because the way does apps. I've been\Ncollecting data. If you know, if there's
Dialogue: 0,0:48:52.01,0:48:56.98,Default,,0000,0000,0000,,one thing to take out of this of this dog\Nis that, it's been nothing short of
Dialogue: 0,0:48:56.98,0:49:02.37,Default,,0000,0000,0000,,horrifying, really. That data is being\Ncollected before and shared before you
Dialogue: 0,0:49:02.37,0:49:06.32,Default,,0000,0000,0000,,even get your consent to anything. I\Nwouldn't trust any of these private
Dialogue: 0,0:49:06.32,0:49:16.10,Default,,0000,0000,0000,,companies to really be the ones carrying\Nwell taking part in in in medical research
Dialogue: 0,0:49:16.10,0:49:22.75,Default,,0000,0000,0000,,or on those. So I agree with you that\Nthere is a need for better and more data
Dialogue: 0,0:49:22.75,0:49:28.86,Default,,0000,0000,0000,,on women's health. But I don't think. I\Ndon't think any of these actors so far
Dialogue: 0,0:49:28.86,0:49:33.90,Default,,0000,0000,0000,,have proved to be trusted on this issue.\NHerald: Microphone 2.
Dialogue: 0,0:49:33.90,0:49:37.01,Default,,0000,0000,0000,,Mic 2: Yeah. Thank you for this great\Ntalk. Um. Short question. What do you
Dialogue: 0,0:49:37.01,0:49:42.28,Default,,0000,0000,0000,,think is the rationale of, uh, this\Nmenstruation apps to integrate the
Dialogue: 0,0:49:42.28,0:49:46.47,Default,,0000,0000,0000,,Facebook SDK if they don't get money from\NFacebook? OK, uh. Being able to
Dialogue: 0,0:49:46.47,0:49:54.16,Default,,0000,0000,0000,,commercialize and this data.\NChris: Good question. Um, it could be a
Dialogue: 0,0:49:54.16,0:50:00.91,Default,,0000,0000,0000,,mix of things. So sometimes it's literally\Nthe the the developers literally just have
Dialogue: 0,0:50:00.91,0:50:05.11,Default,,0000,0000,0000,,this as part of their tool chain their\Nworkflow when they're developing apps. I
Dialogue: 0,0:50:05.11,0:50:08.28,Default,,0000,0000,0000,,don't necessarily know about these two\Npeer trackers where other apps are
Dialogue: 0,0:50:08.28,0:50:14.08,Default,,0000,0000,0000,,developed by these companies. But, uh, in\Nour in our previous work, which I
Dialogue: 0,0:50:14.08,0:50:18.63,Default,,0000,0000,0000,,presented last year, we find that some\Ncompanies just produce a load of apps and
Dialogue: 0,0:50:18.63,0:50:22.55,Default,,0000,0000,0000,,they just use the same tool chain every\Ntime. That includes by default. The
Dialogue: 0,0:50:22.55,0:50:29.55,Default,,0000,0000,0000,,Facebook SDK is part of a tool chain. Uh,\Nsome of them are like included for what I
Dialogue: 0,0:50:29.55,0:50:34.27,Default,,0000,0000,0000,,would regard as genuine purposes. Like\Nthey want their users to share something
Dialogue: 0,0:50:34.27,0:50:37.78,Default,,0000,0000,0000,,or they want their users to be able to log\Nin with Facebook and those cases, they
Dialogue: 0,0:50:37.78,0:50:42.21,Default,,0000,0000,0000,,included, for what would be regarded a\Nlegitimate reason below them. Just don't
Dialogue: 0,0:50:42.21,0:50:47.76,Default,,0000,0000,0000,,ever actually they haven't integrated it\Ndoes appearance and they don't ever really
Dialogue: 0,0:50:47.76,0:50:52.07,Default,,0000,0000,0000,,use anything of it other than that. Mean\Nthat there are a lot of developers simply
Dialogue: 0,0:50:52.07,0:51:02.46,Default,,0000,0000,0000,,quite unaware of the default state is\Nverbose and how it sends data to Facebook.
Dialogue: 0,0:51:02.46,0:51:06.22,Default,,0000,0000,0000,,Herald: Yeah. Maybe we be close with one\Nlast question from me. Um, it doesn't it's
Dialogue: 0,0:51:06.22,0:51:12.12,Default,,0000,0000,0000,,usually a bunch of ups. How many of them\Ndo certificate pinning? Uh, we see this as a
Dialogue: 0,0:51:12.12,0:51:16.92,Default,,0000,0000,0000,,widespread policy or...\NChris: Are they just not really. Yet. I
Dialogue: 0,0:51:16.92,0:51:21.93,Default,,0000,0000,0000,,would have a problem doing an analysis\Nwhere stuff could've been pinned. You say
Dialogue: 0,0:51:21.93,0:51:28.71,Default,,0000,0000,0000,,TLS 1.3 is proven to be\Nmore problematic than pinning. Uh, yeah.
Dialogue: 0,0:51:28.71,0:51:32.41,Default,,0000,0000,0000,,Herald: Ok, well, thank you so much. And,\Nuh. Yeah.
Dialogue: 0,0:51:32.41,0:51:40.52,Default,,0000,0000,0000,,{\i1}Applause{\i0}
Dialogue: 0,0:51:40.52,0:51:44.20,Default,,0000,0000,0000,,{\i1}36C3 Postroll music{\i0}
Dialogue: 0,0:51:44.20,0:52:08.00,Default,,0000,0000,0000,,Subtitles created by c3subtitles.de\Nin the year 2020. Join, and help us!