Hacker News Reader

Comment by alfredallan1 | Hacker News Reader

I’m not sure I entirely follow the point of the complaints. The complainants seem to have a problem with the fact that they’re asked to either agree with the ToS or not use the product/service. How’s that illegal/unfair? No business is obliged to allow all and sundry to use their product/service - regardless of whether they abide by the rules. GDPR doesn’t mandate that, does it?

While I of course fully support the protection of consumer data/privacy, the companies also have a right to decide whether they want non-compliant users using their service. They’re not running something like public transport, which must, by rights, be available to all.

Are these complaints in fact valid under GDPR?

View on HN

There are several lawful bases for processing personal data. Consent is one of them. The companies don't have to rely on consent. But if they do that consent must be freely given.

This point in GDPR could be clearer, and I hope the national regulators provide better advice.

Also, lots of complainants don't understand GDPR and so we will see a bunch of complaints based on incorrect understanding of GDPR.

https://gdpr-info.eu/art-7-gdpr/

> When assessing whether consent is freely given, utmost account shall be taken of whether, inter alia, the performance of a contract, including the provision of a service, is conditional on consent to the processing of personal data that is not necessary for the performance of that contract.

>Also, lots of complainants don't understand GDPR and so we will see a bunch of complaints based on incorrect understanding of GDPR.

Indeed, probably lots of entitled people who believe Facebook (though I’m no fan, and haven’t used it in years) ought to behave as a charity or govt-funded entity.

>https://gdpr-info.eu/art-7-gdpr/

Thanks for that (I’ll be reading through the link more fully).

>When assessing whether consent is freely given, utmost account shall be taken of whether, inter alia, the performance of a contract, including the provision of a service, is conditional on consent to the processing of personal data that is not necessary for the performance of that contract.

So, the company needs to phrase/frame/present the “service” carefully enough - to take the example of FB again, the service they offer could be phrased as "a means to communicate with and see updates from your connections, together with ads that have been targeted based on your profile details”. Naturally a contract to execute such a service cannot be fulfilled without them processing personal data, and this is the only service they offer. Does this understanding seem correct?

> No business is obliged to allow all and sundry to use their product/service - regardless of whether they abide by the rules.

If they are operating in the EU? they do actually. You can't just put whatever you want in your ToS/EULA and throw your hands up.

We should already be used to this. We're used to seeing the phrase "Your statutory rights remain unaffected". A site's ToS, a store's policies, etc do not override your rights available by law.

If a shop says "no returns", what they actually mean is "changing your mind or deciding you don't like it isn't a valid reason". But you can still return something if it's faulty, isn't as described, or isn't fit for purpose - because they're your legal rights, phooey to the store's return policies.

It's my understanding (which could be wrong, is constantly evolving, and desperately needs test cases) that being able opt out of unnecessary PII collection or processing is a legal right, and phooey to the ToS that claims otherwise.

In principle...

In practice, I think that's yet to be determined. GDPR is new. These rights are new. "If it's faulty" is easy for lawyers to determine. "Fit for purpose" ...isn't that hard, and the lawyers have had time to practice. "Unnecessary PII collection or processing" ... the lawyers are currently 23 wikipedia links deep^ trying to identify the aristotelian truth of the matter, presumably in order to make necessary amendments to the aforementioned contract.

Shops putting signs on walls is just not comparable to this, in practice. Normal people are signing dozens of pseudo-contracts they don't understand each day. That has been accelerating and gdpr has accelerated it further.

>It's my understanding (which could be wrong, is constantly evolving, and desperately needs test cases) that being able opt out of unnecessary PII collection or processing is a legal right, and phooey to the ToS that claims otherwise.

That’s actually a great point. Collecting PII is NOT illegal, as far as I can see. Nor can it be illegal. Same for personalized ads.

Consider a real estate agent. To show you the right kinds of houses to buy and finally make the contract, they have to have enough PII about you. If they know nothing about you, they’ll end up showing 1-bedroom studios for a family of five and a three bedroom suite for a bachelor living alone. Similar arguments can be made for all kinds of service providers.

Now onto the issue of FB, the only service they offer is “communication tools with built in personalized ads”. PII is necesary to provide that service.

well, I'm cagey with "unnecessary" because I honestly think it's going to take a good few cases in front of judges until we properly understand where that lays. Which will be fun.

I'm also curious where other points will land in reality. eg, if I ask Apple to remove all data they have about me, will that also remove me from your address book? If I ask Google to remove all data they have about me, will that also remove any emails you've received from me?

In this light, I'm happy to see the first complaints arrive against Google and Facebook, as they're exactly the examples we all have in mind, and we need to see how they'll play out in reality.

(Albeit, I also trust the European legal systems won't commonly wield the maximum possible fines, but rather seek compliance. The scary big numbers, the source of so much FUD, should be reserved for wilful disregard after this step. "Walk softly but carry a big stick" style.

Yes, indeed. It will take a few cases to see what reality is. Though I am equally sure companies which make billions off personal data will leave no stone unturned to argue that their use case is legit.

Their case will be helped by the fact that they can simply have the user “consent” to stuff. And since none of those are “critical” services, they will find ways to prove that user consent was actually voluntary, and not coerced.

So I think you’re right that only the most egregious violations will have real trouble.

The judges might also want to demonstrate to everybody (including their own system) that the big stick can in fact be wielded to real effect. If people start beliving that the stick is really just a prop, no point carrying it, is there…

> Collecting PII is NOT illegal, as far as I can see. Nor can it be illegal.

Article 6 [0] is phrased negatively, making collecting PII illegal unless x or y. These points cover all the use cases the lawmakers deemed valid; a real-estate agent may collect PII because of 6.1b (taking steps to enter a contract at the request of the data subject). Should a new, possibly valid reason to collect PII come up it would first need to be checked and then added to the list.

Since most websites and online services do not aim to form a contract nor fit points c-f, they have to obtain consent by the data subject (6.1a) to make collecting PII legal.

[0] https://gdpr-info.eu/art-6-gdpr/

Since most websites and online services do not aim to form a contract nor fit points c-f, they have to obtain consent by the data subject (6.1a) to make collecting PII legal.

Many common purposes for data processing might come under point f (legitimate interests) depending on the circumstances. Analytics, marketing, monitoring to check for fraud or other abuses are just a few examples.

Doesn’t “consent” (point a) supersede all the rest? Looking at how people have been conditioned to so easily click on "I agree", that will be the most obvious thing to pursue for most companies.

You can do more or less anything with the subject's consent, but one of the big changes under the GDPR is that subjects can always withdraw that consent later and then exercise their right to erasure if there is no other lawful basis for that processing.

True, but the custom isn't that we sign seven contracts at lunchtime, as we browse shops in a mall & buy a sandwich . That's new, for a regular person "seeing" gdpr.

We'll see if this difference is legally superficial or material, as this all plays out.

There was a lot of MOP interest in gdpr and it's friends (like the US Congress' FB stuff). Politicians got attention for their rhetoric, as they legislated. Journalists delivered coverage and opinion, including a lot of moralising too. Social media was abuzz.^ They never told the average MOP that gdpr would be a bunch of contracts she'd be signing.

From the perspective of corporations.. the party complying & implementing gdpr... they've mostly interpreted it as "Rules About The Contract Our Customers Must Sign". Their lawyers assure them this is best.

I think this was a piece of legislation that our legislative systems are particularly ill-equipped to deal with.

^Relatively to the standard interest in laws.

Fun fact: if you buy a bread at a German bakery, you‘re entering into not one, but three distinct contracts.

It‘s just that we usually don‘t notice it because it‘s all concludent action, not a signature.

Yes, it‘s strange, but it is the holy cow of German contract law.

I suspected this was the case, but had no idea of the search terms. It strikes me I do enter several contracts per day, offline, almost always simply implied by actions. And we have consumer protection acts to guard the edges of these implied contracts.

I think what the web needs right now is to figure out what the implied contract is, and save the modals for those out of the ordinary.

Pardon my apparent impertinence in putting it this way - but isn’t the implicit contract rather obvious?

In the digital/information age, personal data is the new currency. You pay for online services with that currency. If your data (money) is more valuable to you than the value you’d derive from use of a certain service, then you don’t pay for it (and by rights, don’t get to use it). I think my money is too valuable for me to spend it on bottled water, so I drink tap water. I think my personal data is too valuable for what I’d get from using FB, so I dont use it (there’s 20+ very widely used apps to “stay in touch” and “communicate” with people).

Eventually (even a couple decades maybe), it will become a question of getting the most bang for your buck (data). If two companies offer the same sort of service, but no. 1 needs all personal info all the time (like FB today), and no. 2 only wants your location twice a day, no. 2 will be cheaper and win out. When no. 2’s executives get greedy and ask for your location 4 times a day, Reddit will squeal.

No, the implicit contract is far from obvious. Yes, everyone knows Facebook shows you advertisements in return for using their service; that personal information is used for that purpose, the extent of personal information collected, and what else it can be used for and with whom it may be shared, is far from clear even to technically-inclined individuals, much less the average user.

There's a reason Facebook has you agree to a 10,000 words ToS and related documents, and it's not because the contract here is blindingly obvious. It would be another thing if the type and extent of their data collection was clearly spelled out and delineated so you knew exactly what you're giving up; but Facebook explicitly wants the ability to collect anything and everything and use it for whatever. Under those conditions, what kind of information they collect about you and what other data it allows them to infer is completely opaque. ("It's just an indoor photo me and a friend, surely Googlebook can't tell where this was taken?" https://nakedsecurity.sophos.com/2016/02/29/google-knows-whe...)

You’re right that what they collect and what what they use it for should be specified more clearly. I suspect if one carefully read through the 10K word ToS, it covers just about everything they do. If it hadn't historically, it will have, since GDPR.

Despite the whole media stink about Cambridge Analytica, usage figures aren’t significantly different. This leads me to believe the average user will simply agree to whatever is put in front of them so be able to use checkout pics and read newsfeeds or play farmville or whatever else it is that people do on FB.

What the EU could mandate is a max 1000 word ToS, that the more concerned user without legal training could actually grok. The GDPR just makes sure companies cover themselves legally - not that hard, if they can get away with a humongous ToS which almost no one will ever read.

That's not fun at all :)

I agree. Still, I think that there is a difference between implied contract/obligations and an actual piece of paper with your signature (or checkbox) on it.

GDPR has requirements about data protection, right to forget, pseudo-anonymity, etc. But as long as the ToS doesn’t violate the rest of GDPR (i.e. things like you allow us to store your passwords in plain text), are there any actual restrictions are imposed on ToS?

Now, for the complainants complaints to be valid, there have to be things in the ToS that are against GDPR. As far as one can see, there isn’t. BigCos have been careful enough to not step on the toes of EU regulations. But what part of the GDPR gives the EU the right to force a company to offer its service to someone who doesn’t accept the “house-rules”?

Sorry if I seem argumentative, I’m not, I’m just trying to wrap my own head around this thing.

The problem is not with "accept ToS or leave", but that that ToS demand things for the company the complainants argue are not legal to blanket demand in ToS, but have to be optional.

The GDPR says no such thing, does it? My GDPR skills are limited by a reasonably good read of wiki and other articles, not the actual text. It states “Unless a data subject has provided explicit consent to data processing for one or more purposes, personal data may not be processed”. It has other requirements about the company being “responsible”, protecting by design, pseudo-anon, right to forget, etc.

As far as I understood, it doesn’t mandate the exact nature of the ToS between the service provider and user, leaving that to be a free-will agreement between two parties.

Consent is only one basis for processing, and comes with high requirements.

Consent is required to be "freely given", which isn't the case if it is required. E.g. Article 7:

Or in Recital 43:

> Consent is presumed not to be freely given [...] if the performance of a contract, including the provision of a service, is dependent on the consent despite such consent not being necessary for such performance.

Thus, where they do give "consent" as the basis on which they process data as part of Terms of Service the user has to accept, that's highly critical if they do not query for that consent independently, and in a way the user clearly can deny. You can't require consent in something the user has to accept to use your service.

(You can find the full complaint against Facebook here: https://noyb.eu/wp-content/uploads/2018/05/complaint-faceboo...)

>Consent is presumed not to be freely given [...] if the performance of a contract, including the provision of a service, is dependent on the consent despite such consent not being necessary for such performance.

Yes, indeed. That just leaves the matter to specifying the exact nature of the contract. So the service being provisioned can be “access to the tools to connect with one’s friends and such tools have personalized ads built into them”. Does my approach misinterpret things? I don’t think anything is specified about what kinds of contracts can or cannot be offered. And in a reasonably free market, it cannot. And of course, it should not - hardly any of the biggest internet companies could make money without ads.

Now if FB or Google, or another company were to be regulated as a utility company, then yes, the nature of their contracts can also be regulated.

They could go a step further, and simply ban “all personalized ads” - I’m not sure how feasible that is, but that would give the EU the right to regulate the kinds of ads shown by FB and ergo, the kind of data collected. But personalized ads is a immensely vague term that could even catch tv and radio ads (targeting based on demographics, and time of day the show is being broadcast, etc.) in the dragnet.

From my understanding that's part of the point of those complaints: a) force Facebook to clearly use that argument if they want to rely on it (GDPR requires them to be explicit about base of processing, and right now they're not) and b) regulators and eventually courts to figure out where those lines are, and get Facebook to adjust accordingly. It's certainly possible that they find that advertising-financed offerings can justify some processing by the need to make money, but that needs to be determined. E.g. I could imagine that they'll be blocked from using protected categories of data that way without explicit opt-in, but will be allowed some matching on others.

To be a bit glib, if Ads are a core part of the offering, surely the facebook.com front page should mention them?

Ah, that’s a good idea. The EU could/should mandate specifying very clearly what precisely the service is.

Of course, like another commenter mentioned about face recognition and image processing, that is hardly relevant to most kinds of advertizing.

The question is one of legal basis and also of specifying explicitly what is being done with the data.

That leads to the next question - how will things be verified? Will an entity get to audit FB algorithms in toto to sign off that they’re secretly processing facial recognition data?…that’ll be the real challenge.

> It states “Unless a data subject has provided explicit consent to data processing for one or more purposes, personal data may not be processed”.

It doesn't really say that. There are six lawful reasons to process data. User consent is only one of them. The one that's useful for most companies is "legitimate interest".

https://gdpr-info.eu/art-6-gdpr/

See, for example, advice from the UK regulator about potentially using legitimate interest for marketing (something that many people are saying needs explicit consent): https://ico.org.uk/for-organisations/guide-to-the-general-da...

>There are six lawful reasons to process data. User consent is only one of them.

Yes indeed, there is a list of reasons. I pick the one based on consent, because it seems the most relevant for a company. Process the users’ data, but get their permission first.

>The one that's useful for most companies is "legitimate interest”.

I would have thought that one was a bit nebulous. No one has given a hard and fast definition of that term. And rather than risk a court’s interpretation or benefit of the doubt, I figured it might be safer to rely on the consent approach.

Consent is something you can easily get, but if a person declines to consent, you may not shut him out.

But if you can argue with legitimate interest (which is more burdensome), the person cannot decline and still use the service.

Choose: either easy basis and accepting that you might not get your way, or difficult basis and getting your way.

If the service I offer is “communication tools with in built personal ads”, or “latest news reports with personal ads on the side” it is not possible to offer those services without access to personal info.

And if the service you offer is „communication tool with the ability to sell weapons online“, then the authorities are out of luck, because you have cleverly phrased your mission statement?

That‘s not how it works.

Oh no. The trouble with “chat app with weapons sales feature” is that it is illegal to buy or sell weapons (online or offline) without specific licenses.

Now If I made a “communication tool” exclusively for registered gun dealers and people with valid (and verified) hunting licenses, it would in fact be perfectly legal. Similarly, if the tool were for use by government agencies and defense contractors or registered weapons manufacturers, it would also be legit.

In the context of “chat tool with personalized ads on the side” - there is nothing specifically illegal about “personalized ads”. Is there? (As long as there is user consent of course). So what’s the specific legal hurdle? Which paragraph of the GDPR prevents this?

> How’s that illegal/unfair?

Because online privacy is now an inalienable right. Users can't sign it away permanently, and companies can't ask for it in return for something.

> No business is obliged to allow all and sundry to use their product/service

No, but they can't discriminate with regards to submission to online tracking.

> the companies also have a right to decide whether they want non-compliant users using their service.

In the EU there are no "non-compliant" users when it comes to agreeing to tracking; there are only non-compliant publishers. These publishers are free to shut down EU operations and block EU ip-addresses, and I wish they would instead of these now common "opt out of any of our 50+ 'partners' individually through this terrible interface" shenanigans, designed to make it cumbersome to exercise one's right not to be tracked online.

> Are these complaints in fact valid under GDPR?

I believe so, and if not, they should be.

Reply on news.ycombinator.com