News for hackers. Stuff that matters.
Re:Dark patterns in GDPR consent boxes by andrepd on Tuesday July 07, 2020 @08:10AM (#23758332)
This interface should be all handled by the browser. It was a mistake not to do this. It makes no sense that blocking cookies should be done via inconststent and dubious interfaces implemented by the websites themselves. Much better that the law would specify minimum standards for browser interfaces to make those choices (which btw would be trivially implemented, vs forcing the sites to spend time and money to implement a shitty js pop-up). Plus you are assured that cookies are really blocked.

As for server-side fingerprinting, the browser, upon the user's choice, would send the allow/deny information to the website, which is forced by law to honor it (as it is today).

Re:Dark patterns in GDPR consent boxes by wjdp on Tuesday July 07, 2020 @12:03PM (#23760577)
How can a browser differentiate between a first-party cookie needed for login and a first-party cookie for tracking?

It's legal to use cookies for behaviour such as login which is necessary but you need consent for tracking.

If a browser blocked all cookies until the user turned them on you'd have the choice of "no login" or "login works, but so does all the tracking".

Not saying the current state of affairs is good, it's awful.

Re:Dark patterns in GDPR consent boxes by _puk on Tuesday July 07, 2020 @01:24PM (#23761589)
If we're going down the route of regulation, and browser control, would it be a step too far to require standardised metadata in a cookie?

That way, each cookie could describe itself as login, tracking, optional functionality etc.

You can then penalise on cookies that purposefully violate this, and allow the user to centrally opt in or out of each type.

Re:Dark patterns in GDPR consent boxes by wjdp on Wednesday July 08, 2020 @06:30AM (#23768575)
Aha, place trust in the site to label their cookies correctly?

I see you've covered that with penalising sites who mislabel. Who maintains this list? If it's the browser vendor remember which company owns the largest share in this market.

Re:Dark patterns in GDPR consent boxes by saagarjha on Tuesday July 07, 2020 @02:23PM (#23762358)
The evil bit is currently little used, I suggest repurposing that: https://en.wikipedia.org/wiki/Evil_bit
Re:Dark patterns in GDPR consent boxes by qwerty456127 on Wednesday July 08, 2020 @05:35PM (#23774499)
"Evil" is subjective and too broad (most of it has already been outlawed anyway and that works: e.g. a competing business is going to think twice before DDoSing you as they know they are doomed if that gets discovered and proven). Surveillance, however, is a much more specific thing. Stalking people already is illegal for people, it should also be made illegal for companies.
Re:Dark patterns in GDPR consent boxes by tgsovlerkhgsel on Tuesday July 07, 2020 @04:53PM (#23763943)
One simple factor could be "does the user have a password (or WebAuthN credential) stored for this web site".

Otherwise have a way for the web site to trigger a browser-controled consent UI. This would be a one-shot thing and clicking "no" would trigger a spam signal. Too many of these and the web site loses its "ask for cookies" permission.

Cookies for anything not trusted have a lifetime of "until the tab is closed".

Re:Dark patterns in GDPR consent boxes by qwerty456127 on Tuesday July 07, 2020 @07:21PM (#23765380)
> This would be a one-shot thing and clicking "no" would trigger a spam signal. Too many of these and the web site loses its "ask for cookies" permission.

Don't be too rough on them. You should look for a viable and sustainable solution, not a radical GtFO doomed to end here.

> Cookies for anything not trusted have a lifetime of "until the tab is closed".

"until the tab is closed" policy isn't convenient even for the user. I tried it and reverted to "until the browser is closed" quickly.

Re:Dark patterns in GDPR consent boxes by joekrill on Tuesday July 07, 2020 @08:20AM (#23758404)
This is a very naive response. Rarely is anything "trivially implemented". And if it's so trivially implemented, why haven't we seen a browser vendor implement something already? Websites have been circumventing attempts to prevent fingerprinting for a long, long time. To act like this is some simple task is hugely misleading.
Re:Dark patterns in GDPR consent boxes by andrepd on Tuesday July 07, 2020 @10:00AM (#23759350)
I think you may have misunderstood my comment. I meant that the "consent forms" should be a browser-side feature, consistent across websites and such that things like cookie blocking are guaranteed to be honoured. This is trivial.

Things like server-side fingerprinting must be enforced through legislation, e.g. it's against the law to track users if they've selected not to in the consent form.

Re:Dark patterns in GDPR consent boxes by gameswithgo on Tuesday July 07, 2020 @08:36AM (#23758534)
Old man programmer here. I also think it is trivially implemented. What is hard is having the world agree to and craft such a standard.
Re:Dark patterns in GDPR consent boxes by kergonath on Tuesday July 07, 2020 @11:23AM (#23760168)
Young person programmer here. There are a couple of privacy-oriented browsers. If it were really trivial, it’d have been implemented already. Just look at the cat and mouse game between Apple and advertisers around Safari, and the new fingerprinting techniques that show up every now and then. Not much is trivial in web browsers.

There is zero need for standards, which won’t happen anyway because of the power Google has with Chrome.

Re:Dark patterns in GDPR consent boxes by _jal on Tuesday July 07, 2020 @12:19PM (#23760739)
I think you didn't understand the proposition. Browser authors cannot unilaterally implement a proposal that has server-side and legislative components.
Re:Dark patterns in GDPR consent boxes by kergonath on Tuesday July 07, 2020 @06:12PM (#23764723)
We cannot rely on server-side protections, because the people who control the servers are often acting against us. Standards won't change that; the linked article is about how website go against a law. We know what change a standard and gentleman's agreement can bring. We've seen it with Do not Track.

The only protections we can have right now are client-side.

Re:Dark patterns in GDPR consent boxes by sigwinch28 on Tuesday July 07, 2020 @09:45AM (#23759171)
1) Implement the Do-Not-Track header on the client side

2) Pass a law making it illegal for a company with business in the EU to not honor a Do-Not-Track header, with a transition period of several years.

Step 1 is technically very easy. Step 2 is legislative.

Re:Dark patterns in GDPR consent boxes by Guest0918231 on Tuesday July 07, 2020 @12:26PM (#23760800)
I agree with the browser implementation. Automatically blocking the cookies will likely not work because sites frequently combine tracking and required cookies.

1. When the user starts the browser for the first time, ask if they want to allow tracking cookies on all websites.

2. When the user visits a website, pass that tracking answer as true or false. Firefox and Chrome have buttons beside the URL already for 'Site Settings'. Allow users to override their global tracking setting with a per site settings there.

This would be infinitely better than the mess we have now, where every website gives us a pop-up with an intentionally confusing interface. Why can't I say 'No' to tracking once? Why do I need to do it countless times a day, each time navigating a new and confusing interface?

Re:Dark patterns in GDPR consent boxes by chriswarbo on Tuesday July 07, 2020 @12:55PM (#23761160)
> When the user starts the browser for the first time, ask if they want to allow tracking cookies on all websites.

Would it be legal/ethical to allow automated pre-commitment to all terms and conditions that nefarious sites may choose to scatter around their pages, many of which won't have been written until after the user had ticked this "agree to everything" box?

> every website gives us a pop-up with an intentionally confusing interface

Any site doing this is breaking the law. Report them please.

> Why can't I say 'No' to tracking once?

Because sites which track you don't want it. After all, they're the ones who invented "cookie banners"; and they could choose to get rid of them by just, you know, not tracking people. Yet they don't.

Re:Dark patterns in GDPR consent boxes by maccard on Tuesday July 07, 2020 @03:14PM (#23762931)
> Any site doing this is breaking the law. Report them please.

To who? What do I say? The issue with GDPR is that it's for all intents and purposes unpoliced and unpoliceable unless you happen to have sway with a local regulatory body.

I live in the UK, and ICO are toothless. Ive filed multiple complaints - inability to opt out, misuse of PII for advertising purposes, and each time have received a cookie cutter response telling me to report it to the company and respond to ICO if it's not to my satisfaction. That was the last I heard of every complaint, despite me following up.

Re:Dark patterns in GDPR consent boxes by chriswarbo on Wednesday July 08, 2020 @11:57AM (#23771134)
It's a shame; but good on you for trying anyway. It's similar to how I treat my vote: not much use on the grand scale of an election; but I need to use it, to avoid undermining any political complaints I have. It's the entry fee to get in the door. What can we do once inside? I'm not sure.
Re:Dark patterns in GDPR consent boxes by maccard on Thursday July 09, 2020 @07:20AM (#23778966)
Thanks. I really felt deflated on the second complaint. Amazon sent a voucher for a video game from [email protected], which was clearly a marketing ploy. I contacted amazon, they said "oh we're sorry you feel that we misused your email", and I contacted ICO, got the cookie cutter response, replied with details, transcripts from amazon, and never heard back.
Re:Dark patterns in GDPR consent boxes by Guest0918231 on Tuesday July 07, 2020 @02:22PM (#23762353)
> Would it be legal/ethical to allow automated pre-commitment to all terms and conditions that nefarious sites may choose to scatter around their pages, many of which won't have been written until after the user had ticked this "agree to everything" box?

Isn't everyone 'agreeing to everything' outside of the GDPR when they visit sites now, without the option of saying 'no'? Isn't everyone covered by GDPR being tricked into 'agreeing to everything' at the moment? Giving users the ability to disable the tracking aspect across all sites with one simple setting seems like a plus here.

> Any site doing this is breaking the law. Report them please.

Has any action been taken against a site for making their opt-out option more complicated than their opt-in option? Why try to regulate how millions of sites prompt users for consent instead of a few browsers?

> Because sites which track you don't want it. After all, they're the ones who invented "cookie banners"; and they could choose to get rid of them by just, you know, not tracking people. Yet they don't.

They didn't invent cookie banners, they added them because they were required by law. The same law could remove cookie banners and require the sites to respect a browser cookie.

Re:Dark patterns in GDPR consent boxes by chriswarbo on Tuesday July 07, 2020 @05:32PM (#23764343)
> Isn't everyone 'agreeing to everything' outside of the GDPR when they visit sites now, without the option of saying 'no'?

If there's no option to refuse consent, then it's not compliant with GDPR. In countries which implement GDPR (mostly EU countries, but I'm the UK and our law implements GDPR but we're no longer an EU member) those sites are breaking the law (that country's implementation of GDPR).

If you're talking about those in countries which don't implement GDPR (or equivalent), then yes; those people are generally not protected by EU law.

> Giving users the ability to disable the tracking aspect across all sites with one simple setting seems like a plus here.

I agree. Again, good luck getting surveillance companies to pay any attention, or prevent them implementing technically-legal workarounds: "Just a moment! We see you've opted out of our advanced partner network. You may be missing out on the latest tailored brand recommendations! Click here to opt back in."

> Has any action been taken against a site for making their opt-out option more complicated than their opt-in option?

Not as far as I'm aware (and I can't see any on https://www.enforcementtracker.com )

> Why try to regulate how millions of sites prompt users for consent instead of a few browsers?

1) Browsers aren't surveillance companies (OK, not all browsers are; e.g. I'm pretty sure lynx isn't meant to be spying on me).

2) GDPR is bigger than any particular technology. It seems reasonable to make some regulation like "The public considers your business model to be exploitative; from now on this requires explicit consent." It seems less reasonable make a regulation like "The technology/product/process/service you provide could potentially be used by others in these specific ways that the public does not favour; you must provide this specific mechanism/option/etc. in case it does get used for that purpose". It's not necessarily a bad idea, but it would be a pretty big ask. Even looking at the current situation, how would this handle apps? What about tracking pixels? What about scanning nearby WiFi network IDs? What about research or hobbyist operating systems? etc.

> They didn't invent cookie banners, they added them because they were required by law.

The intent of the law was to reduce the prevalence of surveillance-based business models. They've always had the option to stop. That would be the preferred option, for those who wrote GDPR, for members of the public who don't want to be tracked, for members of the public annoyed by popups, etc. They chose banners and, to a lesser extent, to gaslight the victims of their surveillance into thinking that GDPR required all these sites chose to break their own UX.

> The same law could remove cookie banners and require the sites to respect a browser cookie.

Again, it would be nice, but I imagine there would be an industry established overnight to provide opt-back-in banners, under whatever guise they can get away with.

Re:Dark patterns in GDPR consent boxes by qwerty456127 on Tuesday July 07, 2020 @04:20PM (#23763556)
> This interface should be all handled by the browser. It was a mistake not to do this. It makes no sense that blocking cookies should be done via inconststent and dubious interfaces implemented by the websites themselves.

I 1000% agree. And usability is not the only reason. I would also name the obvious responsibility delegation/abstraction principle and the original semantic concept of a web site: 1. the functionality every website is meant to have should be implemented at the browser level 2. no website (those using tracking elements included) should be required to have interactive elements (like the consent button) or off-topic texts (like the cookie notification).

Re:Dark patterns in GDPR consent boxes by chriswarbo on Tuesday July 07, 2020 @12:49PM (#23761078)
From the technical side:

We had this ( https://en.wikipedia.org/wiki/P3P ). It didn't catch on.

> It makes no sense that blocking cookies should be done via inconststent and dubious interfaces implemented by the websites themselves.

It also makes no sense that authentication should be done via inconsistent and dubious interfaces implemented by the websites themselves. Browsers have offered login prompts for years, yet they're usually avoided. Some sites hijack scrollbars, so it's not particularly surprising.

From the legal side:

GDPR is not about "cookie banners". It's perfectly fine to use cookies, or any other personal/tracking data, if the user's consent is implied. For example login sessions, shopping carts, game highscores, etc. would break without cookies or something equivalent, and those are features that users want, so we can assume their implicit consent for such cookies. GDPR is perfectly unobtrusive (for end users, at least).

On the other hand, users don't want to be tracked and surveilled. Hence anyone collecting personal data for those reasons cannot assume implicit consent. The choice is simple: either stop doing it, or ask for explicit consent. If you're seeing obnoxious "cookie banners", it's because the operators of those sites would rather mess up their UX with annoying crap, rather than entertain the idea of not being a creepy stalker.

From a pragmatic point of view, making it easier (or even automatic!) for users to give up their rights, in a blanket way to anyone who asks them to, just so unscrupulous corporations don't have to experience negative repercussions of their user-hostile decisions, would seem to defeat the whole point of GDPR.

If you don't want to see "cookie banners", ask the site operators to stop being creeps; that way, they wouldn't need to ask.