Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Consent, GDPR and Google Analytics (cunderwood.dev)
113 points by fenier on Feb 13, 2022 | hide | past | favorite | 96 comments


I’m not the biggest fan of Ben Evans, but he’s right on “privacy fanatism”:

> At a certain point EU privacy regulators will realise: When an EU citizen requests a US internet resource, they provide a US server with their IP address; An IP address is PII; The CIA could record that; Therefore it is illegal to provide any internet resource to anyone in the EU

Source: https://twitter.com/benedictevans/status/1492102034409066504

PS: saying this a German citizen…


Extraterritorial jurisdiction + global nature of the internet causes these problems. We've already seen lots of the reverse: it's illegal to provide gambling to Americans. https://en.wikipedia.org/wiki/United_States_v._Scheinberg

It's also legally difficult to provide bank accounts to Americans: https://www.thelocal.fr/20210924/why-americans-are-finding-i...

Then there was the whole incompatible court orders in re Azure: https://www.theverge.com/2018/4/5/17203630/us-v-microsoft-sc...

Really the only workable outcomes are a global agreement on internet-touching governance (which the US will never accept on principle) or Balkanization. Or I suppose an eternal chasing into new as yet unbanned services.


Thanks, that’s was really insightful. I wonder if global agreements are really unimaginable. There have been quite a few from the old days, e.g. international marine conventions. What do you think?


The last two data exchange agreements between US/EU were overturned. I think it's unlikely at this point unless the USA adjusts some of its surveillance laws.


That gets to the heart of it. Europeans are increasingly uncomfortable using US based services due to how the data is used. It is not inconceivable that there will be multiple Internets based on legal jurisdiction, we already see this with China.


Do you imagine the EU blocking EU citizens from accessing US services? I find that hard to believe. "We're blocking your access to the outside world for your protection" must ring pretty hollow to the people who vote. It works in China because nobody gets a vote.


Extra-territorial laws are one way of achieving the same effect. A logical next-step would be blocking websites from jurisdictions where such extra-territorial laws are unenforceable.

"This website is in a territory not subject to EU regulations governing privacy, security, and content. Do you wish to proceed?"


This would amount to a even worse cookie banner. I hope the EU has learned something from the embarrassment that is cookie banners.


Yes

> Simpler rules on cookies: the cookie provision, which has resulted in an overload of consent requests for internet users, will be streamlined. The new rule will be more user-friendly as browser settings will provide an easy way to accept or refuse tracking cookies and other identifiers. The proposal also clarifies that no consent is needed for non-privacy intrusive cookies that improve internet experience, such as cookies to remember shopping-cart history or to count the number of website visitors.

https://digital-strategy.ec.europa.eu/en/policies/eprivacy-r...


Nope. They are blaming the banners on the "greedy" websites. Not on the broken law.

Kind of like the current US government blaming inflation on... also the companies. Not the money printing.


My website has no banner, and is completely legal. I just use cookies for what they were meant for: As login cookie and to store preferences such as dark mode.

It’s not the EU law that’s broken. It’s intentional that if you want to sell someone’s firstborn you need actual approval and not a clause hidden in the ToS


> the embarrassment that is cookie banners

The idea is great, but shady websites make it so you hate it, so that you hate privacy laws instead of the shady companies tracking your children.


Its a responsibility that should never have been put on websites in the first place. The browser should handle it.

Cookie banners is akin to letting browser extensions create their own permission requests.


It is already a reality that you can't access certain US websites as a European. They block you out because they don't want/don't know if they comply with GDPR. Same effect.


This just demonstrates a level of cowardice on the part of those US-based companies. The extraterritoriality of the GPDR has not been tested.


I remember when the Great Firewall was considered the manifestation of evil by old-time internet users.

It'll be hilarious if European nations decide pursuing GDPR cases is intractable when so many services Europeans use are fully outside the country (and beyond EU enforcement of jurisdiction) and they decide a firewall is necessary to protect their citizens from American surveillance. It would prove China was just ahead of the curve.


Do we then also finally get some of our own internet giants that won't get bought out immediately?


Yup, Section 702 of the FISA act needs to be repealed in order for these judgements to not be relevant.

That's not to mention all of the other, non-legally justified analytics performed by the NSA/CIA etc.


I suspect there's a third outcome within crypto many are quietly pursuing. Looked through the lens of "what if the internet were its own country" a lot of web3 makes a bit more sense.

Or maybe I've read too many Neal Stephenson novels.


That was my "eternal chasing into new as yet unbanned services". The ban wave has largely caught up with big ICOs, but not with "governance tokens" or "NFT based communities".

There's going to be a cycle of "web3 gets big money", "big money fraud in web3", "SEC enforcement against web3", and then the launch of "web4" in 2030.


Crypto can never manage that because the infrastructure running it, the power needed to do that and all the people using it are in countries already.


There’s no issue with that. If a person manually takes their information and mails it to the CIA, that’s also fine.

The issue is if a person visits a resource from a company in the EU, they should be able to expect that that information won’t be passed along to any third party that’s not absolutely necessary. Especially not to foreign governments.

You wouldn’t expect a visit to latimes.com to leak your information to the Chinese Party either.


Maybe I'm just old-school, but I expect when I visit a site I'm leaking some PII (my IP address) to every router between my client and latimes.com to do with as they will.

I wouldn't necessarily expect the CCP to be involved unless Internet routing is having a very bad day, but I'd expect the American government to be involved when hitting an American server.


> Maybe I'm just old-school, but I expect when I visit a site I'm leaking some PII (my IP address) to every router between my client and latimes.com to do with as they will.

Presumably you don't expect the american government to get involved after your request has reached latimes.com though?


Technically, the only thing stopping them is SSH, and that can be handled (as Snowden publicized) by tapping latimes.com's systems on the other side of decryption.

Old-school me would not have expected that to happen. Post-Snowden? It's a definite possibility.


> Old-school me would not have expected that to happen. Post-Snowden? It's a definite possibility.

And... Is that a "favorable" (hope that's the right word, non-native here) thing?


It's neither favorable not disfavorable to me; it just is. Something I file away in the back of my head about how the Internet works right now. Individual uses can be favorable or disfavorable.

If I walk into a store and buy some gum, my face is on their security camera. If the cops are hunting for a murderer, they can pull that camera feed. Is this favorable? shrug. I like my privacy but I also like catching murderers. And I have no expectation of privacy when I step in someone's store; similarly, once I've shipped 1s and 0s to someone else, my expectation is they'll use them as they will, and if I don't like it I'll stop shipping 1s and 0s to them.

This is probably just my American sensibilities talking, but growing up in a culture where I was building a credit score before I knew what that was, I'm not surprised services like Google Analytics are e-gossiping on my preferences (any more than I'd have been surprised if two BBS owners, back in the day, gossiped about their users).


> The issue is if a person visits a resource from a company in the EU

Does it have to be a company in the EU? I thought the GDPR covered any website an EU citizen, resident, or visitor might use, in which case US-based websites might have contradictory obligations to the GDPR and US law.


It depends on Art 3.

https://gdpr-info.eu/art-3-gdpr/

Just because a website exists and may be visited by a EU resident, does not mean that the site automatically has to comply.


It will be hard for a lot of US media making deals with European advertisers to claim they’re not intended for use by European residents, though.


Is that not what 2(a) says- if a service is being provided to an EU data subject, that the regulation applies? At least, that is clearly what the EU seems to be claiming? Sure, if no EU data subject actually accesses the site, it doesn't apply, but the moment one does...


Well I mean think of a store which doesn't accept EU payment or ship to EU addresses, nor target EU residents with Advertising. You'd be hard pressed to say they service EU residents even if the site was able to be visited by EU residents.


No where in Article 3 does it say anything about "targeting" them- it only says if the "service" is "offered", whether or not payment is required. So in broad interpretation, simply serving a webpage to an EU data subject is an act of processing personal data (IP address) of an EU data subject related to offering them a service (the web page itself). That is as long as it doesn't fall into one of the carve outs in Article 2- https://gdpr-info.eu/art-2-gdpr/

It could be argued that such an act "falls outside the scope of Union law;" but that seems to be a matter of contention.


I think we have to look at Recital 23 through 31 they clarify what 'goods and services' mean.

https://gdpr-info.eu/recitals/no-23/


Thank you, that does seem to alleviate some of my concerns as above. I'm not as familiar with EU law, it seems that recitals aren't legally binding equally with the "operative" text. But given the context, it seems unlikely a small blog or web shop that doesn't target EU customers would be in scope.


Recitals play two roles in EU law:

1. They are the legal justification for legislating; The EU is not sovereign, so it cannot legislate of its own accord, the EU must show that the legal powers flow from the treaties. So recitals set out which provisions of the treaties apply, and why the legislators think the law is necessary under them.

2. They are an aid to interpretation; the main body of the law should be read "in the light of" the recitals to understand the legislators' intent and to help ensure there is a consistent application of the law between all of the different courts and tribunals in the EU. These recitals are, of course, not part of the actual legal text and are thus not binding, but they're not inoperative.


They're not legally binding since they're written to be understood as clarifications for the lay-person. Ie, not written in the strict language that courts understand and hence, you might hit edge cases the courts might interprete in ways that you don't expect.


It seems somewhat strange that a company selling a service to EU customers might be in trouble for using Google Fonts in a jurisdiction (e.g. Germany) where there are ways to identify a user by means of IP address [0]; but a weblog that was using Google Fonts might not be, since it's a blog and not a goods-and-services site. Google ends up with the IP address equally in both cases.

[0] https://news.ycombinator.com/item?id=30135264


The US isn't going to enforce GDPR violations, so why does it matter?


Do they have a TikTok?


Well, the obvious responses here are that (1) the law does no such thing, and (2) even if it did, the right target for public concern should be the CIA and the government theoretically controlling its behavior, not the EU.

Even if it came to a point where the EU decided that the only for to keep its citizens safe from US intelligence monitoring were to cut out all access between EU and USA internets, the problem would be the US intelligence framework, not the EU.


Nope. Because legitimate interest kicks in there - you can’t provide the requested service without the IP address. At all.

Turn off Google analytics and you can still provide the service.


Really you should just stop using Google analytics. I know all the data is really fun to look at and can even be useful but it’s a bit like poisoning people just for walking into your business.


> I know all the data is really fun to look at

bit of voyeurism as well


You most definitely can provide a service without having a US IP address. Many services have IPs and servers in the EU instead of the US.


He is not right. Does anyone really think that EU regulators don't know that every request provides the server with an IP address?

Will they start sueing every US company that doesn't comply with GDPR? Of course not. The EU is doing this to build pressure against the US and their surveillance fetish. And it's good that they are, because otherwise, who will?

The US government has proven time and over again that they do not care about their citizens' privacy and straight up lie to their faces. And then there is the CLOUD act, which now starts to affect non-US citizens, too.


IP address by itself is not considered PII (at least not yet).

Context matters. IP address along with other information could be considered PII.


There were rulings finding IP addresses by themselves are already PII[0], because an IP address might be tracked back to a person. E.g. an IP address can potentially be used to go to an ISP and request the subscriber information, and the subscriber information potentially identifies the user of the IP address at a given time, if the subscriber cannot name anybody else who could have reasonably used used the IP address at a given time. Courts found that this abstract risk is enough to qualify IP addresses as PII, as they can potentially identify people indirectly.

The recent German ruling about loading Google Fonts without prior consent explicitly mentioned these rulings and made them a core part of their own conclusions.

[0] The most important ruling is the Breyer ruling (C‑582/14), that found, answering question one, that "dynamic" IP addresses are PII. Further rulings have regularly found that "static" IP addresses are PII, and that you cannot really know what is a "dynamic" and a "static" IP address with reasonable certainty anyway.

"Article 2(a) of Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data must be interpreted as meaning that a dynamic IP address registered by an online media services provider when a person accesses a website that the provider makes accessible to the public constitutes personal data within the meaning of that provision, in relation to that provider, where the latter has the legal means which enable it to identify the data subject with additional data which the internet service provider has about that person."

https://curia.europa.eu/juris/document/document.jsf?text=&do...


These rulings are about personal data, not PII. Please don't confuse the two; it's extremely relevant for IPs.

They are personal data because they are a fact about an identifiable person and thus fall under the GDPR's processing requirements esp. relevantly when transferring to third-parties; but they are not per se PII.


Do we have a ruling that IP address alone isn't PII? I thought the opposite was true.

https://www.fieldfisher.com/en/services/privacy-security-and...


Exactly right.


Yes, taking it literally at the extreme case, the rule is unreasonable.

But Google Analytics is the kind of thing the Law was created to stop, it's not an unreasonable unintended effect.


> ... When an EU citizen requests a US internet resource, they provide a US server with their IP address; An IP address is PII; The CIA could record that; Therefore it is illegal to provide any internet resource to anyone in the EU

Forget that. An EU user visiting an EU site might have their packets routed through an entity outside the EU anyway, without their intent and certainly without their explicit consent.


An IP address is personal data, it’s only PII in combination with other data. Don’t collect the other data if you don’t need to.


> it’s only PII in combination with other data

It’s always PII for static IPs, and together with a timestamp it’s also PII for dynamic IPs...


I think you're confusing personal data and PII.


An IP address can legally identify a person, e.g. in the industry of lawyers sending cease & desist notices (and taking you to court) if you torrent something.

There’s a whole bunch of legal precedent for that in the EU.


Providing the IP address for the communication channel is quite obviously necessary and does not require explicit consent.

https://gdpr-text.com/read/article-49/#para_gdpr-a-49_1_1b

> In the absence of an adequacy decision pursuant to Article 45(3), or of appropriate safeguards pursuant to Article 46, [...] a transfer [...] of personal data to a third country or an international organisation shall take place only on one of the following conditions:

> [...]

> (b) the transfer is necessary for the performance of a contract between the data subject and the controller or the implementation of pre-contractual measures taken at the data subject’s request

> [...]

GDPR does not forbid providing internet resources to EU users, that is simply a lie. All it requires is that data handling happens in the best interest of the user.


True, but storing the IP address server-side for purposes other than serving the HTTP request doesn't fall under (b).

Diagnostic logging (e.g. apache logs) is probably okay as long as the organization can show that these logs are destroyed in a reasonable timeframe, but FAFAIK even that is legally a gray area (in the sense that it isn't explicitly forbidden nor allowed).


Diagnostic logging is ok as long as you have a sane retention policy.


Recent court orders in Germany and France beg to differ.


Opening a communications channel to a third party, e.g. a shady spyware company like Google, requires content, which makes sense.


I don’t remember any case related to accessing first party resources, can you give a link ?



That case is not about accessing first party resources. It was about a German website which (effectively) shared data with a third party provider from a country with no adequate privacy protection.


Well, the other side of the coin is US intelligence agencies monitoring and collecting all traffic, far outside its jurisdiction.


What's next, Twitter geniuses realizing it's also not legal for the CIA to poison people in foreign countries? Supply weapons to militias? Trade narcotics?


That seems a ridiculous interpretation. US companies liable for actions performed by the CIA? Forget GDPR, the entire population of the USA is guilty of war crimes.

If the CIA required web sites to explicitly include a privacy invading snippet, even then it is dubious since it is under duress. And in any case, exactly the sort of stuff you would want laws like GDPR to hinder.


That is false. Businesses outside EU are not bound by GDPR.

The problem is when websites in EU, which are expected to follow GDPR, randomly leak information to businesses outside EU.


> Businesses outside EU are not bound by GDPR.

Business outside the EU, interacting with users in the EU are bound by the GDPR. There might not really be a way (currently) to impose penalties on those businesses for violations, but they are certainly bound by them.


This is such a weird argument. Let's say those things are true (and I think they are reasonably true).

- When an EU citizen requests a US internet resource, they provide a US server with their IP address

- An IP address is PII (well, personal data as far GDPR is concerned, but that's a nitpick)

- The CIA could record that

I don't think how you would go to a conclusion from those that "it is illegal to provide any internet resource to anyone in the EU".

First, it's worth noting that GDPR only applies to companies that specifically target its services at individuals in the EU. Targeting means having an EU office, using an EU domain, providing EU languages such as Polish or allowing payments in EU currencies. If your service makes no effort to provide service specifically for European users there is no need to worry about GDPR - even if you are in the US.

Second, while US services targeting individuals in the EU are legally problematic, this doesn't affect other countries - so I see no reason to say "any" here. For example, a Japanese server is free to provide services at individuals in the EU provided they comply with GDPR as EU has an adequacy decision for Japan.

Also, I would like to point out you can replace US with North Korea in this argument. I think it would be ridiculous to say that if European Union were to disallow sending personal data to North Korea (including IP address) then it would mean that it's illegal to provide any internet resource to anyone in the EU.


> Targeting means having an EU office, using an EU domain, providing EU languages such as Polish or allowing payments in EU currencies.

Nope. There's only a single requirement: having EU users.


GDPR recital 23 says the following:

> In order to ensure that natural persons are not deprived of the protection to which they are entitled under this Regulation, the processing of personal data of data subjects who are in the Union by a controller or a processor not established in the Union should be subject to this Regulation where the processing activities are related to offering goods or services to such data subjects irrespective of whether connected to a payment. In order to determine whether such a controller or processor is offering goods or services to data subjects who are in the Union, it should be ascertained whether it is apparent that the controller or processor envisages offering services to data subjects in one or more Member States in the Union. Whereas the mere accessibility of the controller's, processor's or an intermediary's website in the Union, of an email address or of other contact details, or the use of a language generally used in the third country where the controller is established, is insufficient to ascertain such intention, factors such as the use of a language or a currency generally used in one or more Member States with the possibility of ordering goods and services in that other language, or the mentioning of customers or users who are in the Union, may make it apparent that the controller envisages offering goods or services to data subjects in the Union.


What a tangled web of legal niceties and hypothetical interpretations we've woven here. But the moral arithmetic, toward which European thought is tending, is more brutal and something to which American corporations had better pay serious attention to if they want to keep playing this game.

In general; we hold that "ignorance of law is no excuse", yet in contract law _capacity_ is a key construct, and ignorance very much _does_ play a part. It's not just minors, the mentally-ill, or those incapacitated by drugs or alcohol, discombobulated or bamboozled by other means, who cannot give consent in a contractual relation. In an age where most lawyers and judges, like everyone, mindlessly click-through "agreements" and shrink-wrap EULAs, there's a strong and growing argument to be made that non-expert adults lack genuine capacity to understand technologically mediated relations.

In other words, it's the contract law that underlies this stuff that's coming up for revision, not the surface interpretations. The important matter now is not deliberating whether the letter of the law creates "consent" on this or that occasion, but whether the spirit of the law allows for consent even in principle, given societal standards of digital literacy and the complexity of modern digital interactions.


> In an age where most lawyers and judges, like everyone, mindlessly click-through "agreements" and shrink-wrap EULAs ..

That's an interesting problem. I'm a little disappointed that the route we've gone is having courts decide that this or that bit of EULA isn't binding, but people are still expected to read them and be somehow bound by them. It's kind of difficult for the common man to find out which parts of an EULA are or can be legally binding, so why should they ever be read?

For a while now I've been thinking that EULAs should also be made simple and clear and understandable, kinda like they're forced companies to do now with consent dialogs. No walls of text, no small print, no legalese, and definitely no tons of obviously unenforceable but chilling terms (that the poor reader might think are enforceable).

It does not feel right that people are "agreeing" to something they didn't read anyway (and which if they did, most people wouldn't really understand anyway), and they can only find out what their rights are after the fact.. so maybe we should just say that such agreements are not okay, stop it. It should be easy to understand exactly what you are agreeing to (or possibly we could just have the terms in law and stop this silly game altogether).


Rather than making EULAs simple and clear, how about simply making them void? They don't serve any valid purpose.


For the consumer market I agree. It's not fair to force consumers to accept a one-sided unnegotiable contract.

I think the silly interpretation of copyright law (that any transient copies made by your computer when you install or run a program constitutes copying and thus require the copyright holder's permission) should be undone. You shouldn't need a license to use a program any more than you need a license to listen to a music disk or read a book or watch a movie or painting. I think that interpretation of copyright law is fully based on a technical gotcha and not in line with the spirit of the law.


For a while I tried to actually read EULAs because I wanted to know what I was agreeing to and felt which parts are overly generic gave a good indication of what a company might want to do without explicitly stating it.

But it's just too much and too overwhelming and most of it is just completely cookie cutter legal babble which I can't waste my time with. I recently bought a video game to play for fun and when I first started it I was asked to sign off on a ridiculously long legal text followed by five(!) revised versions of it from various updates. What am I to do? Spend a few evenings reading and comparing those instead of playing the game I'd bought? Realize that I couldn't play the game I'd bought and ask for a refund? Or just scroll through, click to make it go away and enjoy my game without thinking more about it.


This is going to be a hot topic in Germany once the German courts rule it out. Should it say it's illegal to load, we have got loads of work in front of us. One simpler solution that I have seen Zaraz by Cloudflare, which seems to solve this issue. Has anyone had experiences with this?

https://blog.cloudflare.com/keep-analytics-tracking-data-in-...


Still an US corporation, subject to the CLOUD Act.


The author of the blog apparently also wrote about Zaraz in this post:

https://cunderwood.dev/2022/01/30/tag-management-is-no-longe...


Cloudflare is from USA so it’s a quick decision to take.


Good article. We need more of these. GDPR and integrating with 3rd party services can be quite a legal minefield.

I would like to see an article regarding Google Recaptcha. I am currently considering Recaptcha during a login process as a means of protecting against credential stuffing and password brute forcing. But I do not know if this counts as "legitimate interest" as defined by GDPR. And if it doesn't, there really isn't any way to ask for consent in this case, because "denying" consent sidesteps the entire security measure...


> I am currently considering Recaptcha during a login process

Please don't, think about your users. Just use normal rate limiting instead of forcing me to select more god damn street signs.


That makes no sense, one is not a replacement for the other. Spam bots can be slow, and human users can hammer your API. You need both.


You don't. If you want to prevent credential stuffing rate limiting is perfectly fine and that was the only reason EtienneK wanted to use captchas.

Spam bots can be slow yes, but this is about credential stuffing, that can not be done slowly or else it'd take years.


Rate limiting doesn't protect against credential stuffing either -- it doesn't manifest as a brute-force attack, they rarely originate from a single ip/network you could reasonably rate limit against, and even if you could magically rate limit them you've already lost because you let the bots try their stolen credentials in the first place. Your only real defense is to have a system that identifies bots directly to make it so attackers can't automate spamming credentials across a bunch of different sites.

If you let attackers have like 10 attempts per ip per minute and you're a site where a bunch of people have accounts then you're gonna become an instant favorite for testing stolen creds.


I just buy captcha solvers in india or use a bypass like the Recaptcha Text to Speech challenge that regulary gets abused by extensions like Buster. (Yes, you can use google speech to text to solve the text to speech captchas.)

Captchas don't really work, the only things they do is let google track more people and annoy humans. You let your users train their AI for little to no protection agains a willing attacker. I'll admit captchas protect against many scriptkiddies that buy passwords. I'd rather teach my users to use a password manager or support passwordless login than let them click on cars for "security".


Just don't opt in to Google Analytics. I don't.


There's an opt-out, but not an opt-in for Google Analytics. Unless you're referring to simply blocking it via a content blocker script.


Content blocking is also opt-out, in the sense that no browser blocks this by default.


I think you meant opt-in. But it's actually opt-out in the case of Firefox, which has been shipping with Enhanced Tracking Protection enabled by default since 2019. https://blog.mozilla.org/en/products/firefox/firefox-now-ava...


Opt-out means on by default, opt-in is off by default (sorry to state the obvious).

The users above are referring to opting in/out of the tracking rather than the blocking of the tracking - so GA on Firefox by that standard is opt-in, even though in a Firefox specific context the setting is opt-out.

Obviously the EU wouldn't be cool with "but a competing browser with a small market share blocks us by default anyway".


Yeah, I guess it depends on what the opting is for. I could have phrased that better.

What I meant to say: third-party tracking is still opt-out, since you need to actively enable the content blocking to avoid being tracked. The blocking of said content trackers is opt-in of course, but that's not what I was referring to.


Firefox does. Enhanced Tracking Protection, which (I think?) is enabled by default, blocks the GA script.


That's true in one sense. In another sense it refers to not requesting the tracking scripts in the first place (ie. not opting-in). That perspective works better if you think of the browser as a user-agent.

It's not really an importance distinction in the end though.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: