EU tells Meta it can't paywall privacy The EU's Data Protection Board (EDPB) has told large online platforms they should not offer users a binary choice between paying for a service and consenting to their personal data being used to provide targeted advertising. The EDPB opinion [PDF] published yesterday addresses whether it is valid in terms of data protection …
It is an interesting concept nonetheless. I agree that the way they've done it, it's at the very least morally dubious and possibly illegal. If they had done it the other way around, it would probably be just fine. ie they created a subscription service from the ground up and then later offered a "free" service so long as you consent to data collection and ads. Of course, no one will do it that way around because they all want the "free and rapid growth, monetise later" model.
They should be fined on the "wheat/rice and chessboard" principle. On the first day of non-compliance they should be fined one penny/cent (it probably doesn't matter which), the second at two pennies or cents, the third at four pennies/cents and doubling every day of non-compliance until the company realises just how soon they will be made completely bankrupt. That should focus their attention a little bit!
Its well past the time that these enormous companies were fined. No, its time to put the figureheads in prison, after taking their assets. Fines just get passed on to Joe and Josephine Customer.
Problem is, those who make these kind of laws are either shareholders or are in the brown envelope workflow, so it will never happen. Fines it is.
>But a Meta spokesperson said: "[...] subscription for no ads complies with EU laws."
More deliberate obfuscation from Meta. The problem isn't ads. Subscription for no ads has been a thing since before Meta even existed, it's never been a problem, and it's not a problem now. The problem is tracking. Meta needs to get rid of tracking and/or make it a whole lot more user-transparent.
Happy for them to offer a choice of paying for no-ads or being sent ads for a free service. But it must be completely separate from the choice about tracking.
Some people may prefer to allow tracking with their ads, because they will be more useful/interesting ads for them Others may prefer no tracking even though it means the ads are always useless to them. That would is the genuine choice for people.
How do you mean? Targeted ads show me ads for something I just bought, not something I'm looking for. I only need one fridge, one washer, one dryer, ect. Showing me ads for these thing for the next 6 months is deliberately missing the target. At least with random ads there's a slight chance of showing what I am now looking for and there's no need to track me for that.
In feed-based apps like Insta and FB, the entire app is literally driven off tracking your personal data, and feeding you what you engaged with before. Never mind ads, *that is what a feed is*. Insisting on non-personal-data-driven posts, * whether ads or not*, is like legislating for “red courage”. “Red” is an adjective, certainly. But applying it to abstract nouns, that’s a category error. Your legislation is literal gibberish. The words don’t mean anything.
Except that's not how "feed based" works. It doesn't need troves of personal information because 99% of the time, the "feed based" algorithms only feed you more of what you just looked at and "sponsored" stuff that is invariably unrelated to either you or your current feed or anything else you ever looked at. If it was genuinely using the data on you and what you have looked at historically, it would be dropping stuff in your feed that you've looked weeks, months or years in the past where there is new and interesting related stuff. But that never happens. It barely "remembers" what you looked at last week in terms of the feed content.
Of course it needs “troves of personal information”. How else do you think it works? “More of what you just looked at” = a hash-table of the last ten thousand items viewed, weighted by engagement, then a bunch of Principal Component analysis to put you in the correct box, then some K-nearest neighbours stuff.
If you want to pretend that you being in a K-nearest neighbour box with other people who post “Yaas Qween” at Katy Perry, and also have an interest in Jungkook BTS doesn’t make you a 16-22yr old female, who will be interested in other things that 16-22yr old females are interested in…..then go right ahead and be silly about it. But probably not a bad target for awareness ads for tampons. And a lousy target for Saga Holiday ads. But there is literally no “targeted ad-ness” in there. They fail to serve the above person Saga Holiday ads, through exactly the same lines of code that also don’t shove out “80-yr old Japanese great-grandmother bemoans why her grand children don’t visit the family Shinto shrine”, and for the same reason. Different box.
“It would be dropping stuff in your feed that you've looked weeks, months or years in the past where there is new and interesting related stuff”. The algorithm does exactly that. It really absolutely does. All day, every day, 24/7 to hundreds of millions of people. Perhaps if it isn’t doing that for you, you’ve got some ad-blocker blocking everything, and have been hoist by your own petard. Can’t say. But for everyone else, no it does exactly as you request. If you’ve watched a hundred “cat plays with otter” clips, then within minutes of your favourite creator dropping “cat plays with otter again” after six months off, it will be in your feed.
> The algorithm does exactly that. It really absolutely does.
This is contrary to my experience with Facebook and Twitter, and I was not using an adblocker until this month. Over the years, the FB and Twitter algorithms became increasingly worse, hiding posts from people and topics I followed and putting garbage that's never been of intrerest to me front and center in my feed. After months of following recommendations to improve things, by checking and unchecking boxes on long lists of things I might find interesting, I gave up on both Facebook and Xitter for good. I half suspect I was an unwitting part of an A/B experiment designed to see how much feed interest-inversion people could tolerate before cracking.
YouTube, on the other hand, usually has interesting recommendations for me; and I've occasionally even purchased things I first saw advertised on YT.
"Well it’s shit and doesn’t work very well. Same as Amazon recommends"
I have come to the conclusion that "interest based ADnoyance" might indeed by of interest to someone, but usually its not me.
I really dont know whose "interest" this ADnoyance actually displays but after a disgust driven kamikaze attack on lmhosts.sam,locking and loading pihole and each and every ADnoyance blocker i could get my hands on the only device that still pesters me with "ads based on other peoples interest" is the google infested smartphone.
1) You don't need tracking to create a personalized feed. I have a news app where I explicitly gave it a list of topics I'm interested in, and it works very well - better than Google News, actually.
2) Meta doesn't just use your tracking data to create Facebook feeds for yourself, or to show targeted ads to yourself. The data is integrated with data gathered from other users, affecting their own feeds and their own targeted ads. It's exposed to other Meta software, and to third parties. It's used in ethically dubious experiments (the infamous 2014 study). It's used to create shadow profiles of people who don't even have a Meta account. This happens in a completely opaque fashion; and those are just the things Meta has explicitly admitted to, and we don't really know the details.
3) Meta doesn't just gather tracking data from what you engage with on their services. They also provide embeddable trackers, with which they can gather data on what you do on third-party websites. Again, this is fully opaque to anyone who doesn't have advanced technical skills.
Creating a personalized feed does not require any of that crap. Meta could let users explicitly state their preferences, and/or they could create them by getting informed consent and then tracking what they do on their Facebook account alone. All of that would result in an application that works fine, and would be well within the law.
You still don’t get it: “ I have a news app where I explicitly gave it a list of topics I'm interested in, and it works very well [ie this is the product I want]”…… “Meta could let users explicitly state their preferences”.
So…..actually what you want is…a news app, which already exists, and nobody at all is suggesting taking it away from you. Huzzah!
But Facebook, or Instagram, they are not what *you* want. They each do something rather different from your news app. And nobody is forcing you to use them. But, for reasons best known to yourself, what *you* want is to *change* Facebook and Instagram fundamentally until they work exactly the same way as your news app. Why do you want to do that, and what gives you the right to shit all over a different product that you don’t want?
This is really no different from the petrolhead vs MAMIL cyclist beef. Whichever side you’re on, “hey those cyclists they just block up the roads and ignore traffic lights. We’ll fix that. Pass a law requiring all bikes to have four wheels, and a minimum 3.5 litre petrol engine, a driving license and pay road tax. Of course you can do it, I drive a vehicle with a 3.5liter petrol engine, and it works just great”.
"Exactly. Display ads to raise revenue, just not targeted, personal-data-driven ads."
You'd think all this much vaunted AI would be able to target ads based on the users direct Facebook interaction and/or the conversations they are having and target ads related to the topic of conversation by now, without ever needing any personal data or huge trove of data. It has to better than bombarding you with ads for stuff you looked at two weeks ago and bought last week and now have no interest in buying again for a while.
So, Instagram. Forget about ads entirely for the moment, just look at the very concept of “an Instagram feed”. “Your Instagram feed” consists 100% of posts in Instagram database, which have a high match on an individualised basis to other posts which you have engaged with in some way. That’s literally what your feed *is*, it is nothing other than closed-loop completely personalised tracking. Again, forget about ads, what would “your Instagram feed” look like without tracking? It would be…..just the globally highest trending hashtags. Stuff you had no interest in, the Kardashians mixed in with equal measures of Trump rallies, Brazilian soap opera stars, Thai cuddle buddies, Teletubbies, K-Pop, and Cardinal Ratzinger. It would be the same feed for everyone, not individualised, so it would just be a global broadcast channel. This is just not a viable concept, it *makes no sense*.
Nobody is asking you to scroll Instagram. You don’t have to, if you don’t want to. But you do have to accept that for those people that scrolling appeals to, they need to be logged in for Insta to show them any meaningful feed. That is their choice. The whole concept is just gibberish otherwise.
Now consider how ads change this whole picture. An ad is simply a post where the poster has paid to have it displayed in among all the others. That is all it is. Apart from the payment, it goes into exactly the same algorithm mixing pot. So, please explain: *how* could Instagram show non-tracked ads into your feed, which is literally defined as “the place where Instagram puts the targeted stuff for you, based on tracking info”. It’s *gibberish* as a concept. It doesn’t parse semantically. You asking for ads to be “smodge-wollocked”, because that sounds bad and you don’t approve of ads, rather than considering whether “smodge-wollocking” is even a word.
“Subscribe….the accounts you choose to subscribe to”.
And, how would you ever have discovered these accounts to subscribe to?! Presumably you didn’t just stare into the ashes of the flames of your Viking longship, guess a username, and run a search to go to their page to see if you were interested in the content? What actually happens is that a certain % of the scroll slots always contain new-possible-accounts of potential interest. If you engage, you will starting seeing posts from that account; if you don’t you will stop seeing posts from that account.
Again, ads are treated *identically* to non-ads. If you are of the correct demographics to see eg an Insta ad from Hello Fresh, then it will get served to you a few times to gauge interest.. But if you show no interest, it won’t ever show any more ads to you from Hello Fresh. It has discovered that you are a waste-of-time poor match for that advertiser, and that’s the last you will see of them. Same as any music band that you might like, but actually don’t. *That’s targeting*.
All you are doing by “turning off targeting” is a) Ensuring that only a teeny-tiny fraction of the “discovery” items in your feed have any possibility of you being interested in them b) Ensuring that the fraction of “discovery” items in your feed has to approach 90%+ in order for there to be any chance whatsoever that anything ever piques your interest before you delete the app in disgust….and that applies to both normal and ads c) Ensuring that once you have viewed an ad from a company in which you have no interest, refusing to let Meta remember that you don’t want to see the ad again! So they must show it again….and again…..and again. Because remembering what you *didn’t* want to see is literally the definition of targeting.
It would look like Tumblr on the whole which makes suggestions for accounts (not posts) based on tags you have used on your posts, or you can search for tags and add those as separate feeds. Instagram *could* do that but the sparseness of the feed for new users would "put people off". I mean yes, it might, so have a decent onboarding process for new users - it's really not hard. Mastodon feeds are the same as well.
Yep.
That model works (or worked) for Reddit. Let people subscribe to topics they are interested in.
And if they enable tracking, use that information as well to give them a better experience. If you do it well enough maybe some of us would agree to let you track. And others won't because it is a price they are unwilling to pay.
"what would “your Instagram feed” look like without tracking?"
I have some ideas they could try:
1. A list of posts from the list of accounts you selected to follow with a sort box on the top so you can pick between various methods of ordering them (newest, most positively reviewed, most positively reviewed by a lot of people, most reviewed by people in any direction, any other sorters you might find useful).
2. A list like that, but in addition to those posts, you see posts from other accounts that are linked to the accounts you chose to follow. This would use public information like which accounts those people chose to follow.
3. A personalized feed of posts picked by asking you to voluntarily give them some information about what you like so they can search for it. You could push some buttons instructing their algorithm what you like and what you don't like, and various programs could scan for patterns in those then predict whether you'll like certain other posts. They don't need to base that information on what someone else liked if they can use an algorithm on the content. You could deliberately indicate these things instead of collecting and using other pieces of information like how long they think you looked at a thing.
4. An anonymized collection linking views and followed posts, but without including user IDs. It doesn't help when you want to see the full history of someone's actions, but the point is that we don't want them to see a full, or partial, history of someone's actions, nor do we want anyone who breaks in to have the ability to retrieve one.
I don't use Instagram. Do any of these look workable to you?
#1 Falls at the most basic function: how do you ever find accounts to follow in the first place. But also, people want to follow what’s fashionable. Insta, by definition, finds people who are “like” them, and suggests them fashions they want to follow. Your app only works for the consciously unfashionable. As you can appreciate, *by numeric definition*, your app loses to Insta in the free market.
#2 You are just outsourcing the discovery mechanism to “a trusted influencer”. We used to call these “TV channels”. You’ve just reinvented TV, or at best “curated content”. Also, how does “the trusted influencer” discover?
#3 “Asking you to suggest what you’d like”. I think you’ve just reinvented the search box….we used to call that Google. If you want Google, go to google.
"#1 Falls at the most basic function: how do you ever find accounts to follow in the first place."
How do you find sites like El Reg to read? You look around, find someone who recommended them or a link going there, take a look, and decide it's worth your time. Your friends could say "Just saw an interesting post about [topic] on the feed of [user]". You think that sounds nice, so you read the post, and liking that, you follow [user].
"But also, people want to follow what’s fashionable. Insta, by definition, finds people who are “like” them, and suggests them fashions they want to follow."
See options 3 and 4.
"#2 You are just outsourcing the discovery mechanism to “a trusted influencer”. We used to call these “TV channels”."
No, I'm outsourcing it to everyone. All of your friends would be used to create the thing. You could then control certain thresholds. Remarkably like now, but you're only using public information to do it.
"#3 “Asking you to suggest what you’d like”. I think you’ve just reinvented the search box….we used to call that Google."
Automatic searches based on voluntarily given information. Predicted results rather than results tailored to a search term. If I use the search box, I only want to see things relevant to this query I've just entered, not earlier ones. With a predictive algorithm based on content I've marked over many previous days, it's very different than a search box. In fact, nowhere in my proposal was a search box. You could have one, but you could leave it out and still follow my suggestion.
You seem to argue that personalized feeds are impossible without tracking.
That is just not true. Users can explicitly declare their interests. There are real applications, right now, that work just like that, and they work fine. I myself have feeds that don't track me. This thing that you claim to be a logical impossibility, not only it's possible, but it actually exists.
Or, you could also track what the user is doing on the app (and not elsewhere), keep that data isolated from any usage that's not strictly necessary to build that specific user's feed, and obtain informed consent. That would also work, and result in a working automated feed that respects the law.
You are arguing that opaque and pervasive tracking is necessary to build a personalized feed, and that's just not true. Multiple people on this thread have shown alternative models, some of which are actually used and have been used for a while.
“I have feeds that don’t track me”
You mean like RSS? Ok….let me know how many people use RSS vs user base of Instagram. Would you say that it has a one-to-one map with people who browse tech websites, for example? Most normal people think your solution is just a terrible fit for their needs, and would rather wear a frumpy dress from Laura Ashley than even consider doing that.
Honestly, this has such a creepy vibe, it’s astonishing you don’t see it: a bunch of middle-aged male software engineers, want to tell 16-25 year women what to watch. And when they refuse to do so, on the mens terms, the men lobby the EU to make the womens thing illegal. It’s just….ick. Ick, Ick, Ick. No way round it, creepy as hell.
>Honestly, this has such a creepy vibe, it’s astonishing you don’t see it: a bunch of middle-aged male software engineers, want to tell 16-25 year women what to watch.
...how do you get from my proposal of "users should be able to explicitly tell the system what they want" to... that? The direct opposite? You live in a mirror universe or something?
The French data protection outfit (CNIL) said this rancid behaviour was okay, which has led to quite a number of sites saying "allow our eight hundred partners to track you, or cough up".
The massive elephant in the room is that there's actually a third option - refuse cookies, respect non tracking, and serve up non targetted adverts based upon what the site in question is (like cooking stuff at Marmiton might be more relevant than a new sofa...).
But since everybody is a greedy bastard, that option was swept under the carpet and forgotten about.
I will not trust that individual to respect anything he promises.
I trust that he will find a way to track people who have subscribed and opted out of advertising, will get caught and try to brush it off with another "Oops, sorry Your Honor, we made a mistake. Won't happen again, pinky promise."
And then it will happen again.
Him and his sort really need to be hit with multi billion fines, non negotiable, payable now, not allowed 20 years of appeals, and them beng put in prison.
If not, the whole farce will just carry on. ''Yes we were caught, sorry, won't do it again honestly'' does not work. It has been tried and it does not work so go for the only thing that will, asset seizure, immediate billion dollar level fines and jail.
Anything short of that is just the various governments fucking their people over.
"I trust that he will find a way to track people who have subscribed and opted out of advertising"
He already has. Subscribers pay for not *seeing* personalized adds. The wording is always that way and not by coincidence. The tracking continues unabated.
Think of it:
* Technically, with them even excessively tracking non FB / Meta users, it would be rather challenging to stop tracking paying customers when they were not logged in without abolishing all tracking on unrelated websites they place their trackers on.
* Paying for one month, if subscribing to non-tracking would do what is legally required to start with (deleting data not essential for the operation of the service), would get ones track record cleared. Surely FB / Meta won't allow you to get that cheap off the hook if even non FB / Meta members have no single way to opt out of being tracked or have their personal data deleted.
As usual, Facebook wants to gaslight not only its users, but also regulators, into thinking that "advertising" equals "tracking". Newsflash: It does not.
The model of "subscribe or see ads" is nothing new, and a valid way to earn money.
But "subscribe or we'll track you all over the internet and across all your devices" is, obviously, not. That is asking the user for a ransom to comply with the law. But according to Meta, that is somehow different from the local branch of the Legitimate Businessmen ClubTM showing up at you door asking for a donation so they don't do anything illegal...
'a Meta spokesperson said: "Last year, the Court of Justice of the European Union ruled that the subscriptions model is a legally valid way for companies to seek people's consent" '
THe ECJ didn't rule on this -- it was an obiter dictum -- efectively a passing comment -- that any alternative to ads must be 'necessary' and the fee must be 'appropriate. The statement did not thererfore have the force of law to grant anyone the right to seek implied consent via a paywall. The ECJ has now apparently ruled to the contrary.
In one place in the article it says the subscription was to avoid targeted adds which would mean no tracking but it gives room to still show adds.
In another, it's implied that the subscription stops adds being shown, but it says nothing about stopping tracking.
Adds / no adds is one option in and of itself - pay the subscription and no adds, else you get adds. That's fine (unless it's excessive - then along comes methods to block those adds).
BUT tracking is an entirely separate issue: That should always be Consent based.
Oddly, Google are pushing the consent based tracking approach - which is strange considering this is google we're talking about... unless it's them saying 'no one who wants to use our analytics tools can track without consent... except for us!'.
Probably the latter. It's google, after all.
Anyhow: As noted in other comments I've made on this: Paying means you're giving up additional information to Facebook, which negates a chunk of what you think you're paying to avoid. If anything, it just makes things worse as they can now track that payment method, especially if they only offered an 'ad free' experience and said nothing about not tracking you...
The heads of the groups that hand down fines to these companies should have a simple and large clock on the wall that displays FD/m (Facebook Dollars/minute) and use that number to multiply by 3 months worth of time minimum.
They seem far apart, but displaying their sawy in Dollars per Minute helps to put in mind of those tasked with punishing them, how much power they have, how little they have to care about conventional fines and how meaningless conventional fines and legal systems are from the view point of the sentencing body.
And applied in minimums starting at 3 months of yearly time passing, not working hours, but a full quarter of the year in real terms. This means you potentially get 3 of these in a year before your company is running negative and starting to eat into previous years profits. The 4th would be more than a years full profit.
This is the only way they will actually pay attention.
Until such a time, they are minor inconveniences. Meerly a latte in terms of their yearly profits.
That "large clock" inevitably reminds me of the "Stop the Film" segment of Monty Python's Blackmail [YouTube].
The right to shelter from the corporate Eye of Sauron has rightly (and explicitly, now) been added to the list of unalienable rights. Rights that can't be sold for money--like life, liberty, and the pursuit of happiness. Facebook's management will not acknowledge that such rights exist, because they know that everything has its price, and it's usually cheaper than you'd think when it comes to self-respect.
Oh no! Here comes all the hypocritical Reg Freetards demanding that a business gives away everything and makes no money. Not a one of these Freetards has ever ran a business apparently, and mostly likely have mum cook them dinner. The El Reg Freetard has completely ignored the fact that there currently is a third option: Dont use Facebook! But apparently, even though they hate FB with all their hearts not using it is not an option.
Let me fix that for you : Facebook potentiallyaggressively tracks you even if you don't knowingly use them, or even have an account with them
But people are not suggesting Faecesborg (or anyone else) have to give everything away for nothing. Personally, I wouldn't be too worried if they did offer an offer of "pay ${modest} fee, or allow us to monetise your information in the way we explain in simple language here" option. But that's not what Faecesborg are doing, they are deliberately confusing two different issues, and charging significantly more than it's suggested the cost of providing the service would suggest anyway. The "pay ${modest fee} or let us pay for it through more targetted advertising" is also OK provided the fee is reasonable and the details are set out in easy to understand language instead of being hidden behind a gazzillion pages of legalese.
I have no Facebook account, so when occasionally cookies are not blocked, my wife gets my targetted advertising instead of me. So I am interrogated by her along the lines of: Have you (me) been buying wine again... You do not need another Raspberry Pi. We are doing no more jigsaws until Christmas...
So much arguing here about irrelevant things.
There are actual laws about "common carriers". The telephone network and mail system comes under that heading.
Newspapers, TV news, etc do not.
Is Facebook etc more like the telephone network, mail services etc, or is it akin to Newspapers, TV News and other curators and pushers of topics?
Social media platforms can't hide behind "common carrier" laws any longer, as that's most definitely not what they are.
Initial rulings from way back about BBSes and like deemed them "common carriers" so the owners of the platforms weren't responsible for the information on them.
Specialised platforms promoting crimes are not covered by this ruling, BTW.
There is no way that siloing and content pushing platforms are "common carriers" any longer.
“EU law requires that consent is the genuine free will of the user.”
How I wish this were meaningfully implemented on all platforms everywhere.
No more “yes” or “maybe later”.
No more “agree” or “edit settings”
No more “choice they want me to make” in a giant green button and “choice I want to make” in tiny text underneath.
Just yes or no. Or even better, don’t do it and don’t ask me.
And none of those comply with EU regulations. It needs a few sites facing action to persuade the rest that there will be a real cost from ignoring the law. As it is, it's a token attempt which (as you point out) is designed to stop most people exercising their rights.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
