7 ms·

Apple's WebKit team proposes a way ads can be measured while maintaining privacy

125 points by fauigerzigerk 7y ago

dvfjsdhgfv 7y ago
Google has a million ways of tracking us anyway, this is going to be just a minor problem for them they will solve sooner or later.
- bdibs 7y ago
  Does that mean people shouldn’t try to stop it?
  - dvfjsdhgfv 7y ago
    No, I think we should try anyway, but it will be extremely difficult to do and a lot more effort should be put in educating people about what price they pay for "free" product handing over their and others' data into the hands of a single entity.
- StreamBright 7y ago
  Actually there are ways to stop Google like pi-hole for example
  - billyhastings 7y ago
    Facebook ads regularly bypass pi-hole (at least in my experience). However, I disagree entirely with the assumption of "just because they find way we shouldn't try". With that attitude, your ancestors would likely be either Japanese or German.
  - sixothree 7y ago
    I wonder how many people they have working to circumvent pi-hole specifically.
    - adrianN 7y ago
      As roughly zero people who would click an ad use a pihole, I'd guess roughly zero engineers are working on the problem.
    - StreamBright 7y ago
      How would you go about "circumventing" pi-hole? I do not allow DNS requests going towards Google's DNS servers by firewall. The only option would be to use a DoH but it is also blocked on standard ports. The only option is to use a non-Google associated IP with a non-standard port which is not worth it at the scale of Google.
AJRF 7y ago
Forgive the pessimism but isn't this article rewritten every year? It's always along the lines of "ad tech company begs apple not to enable X". Doesn't really seem to be that effective as there is always something to replace the newly blocked method each time.
- paranoidrobot 7y ago
  That would probably be why the subheading is: > Safari tech ready to be ignored by online ad giants like all other privacy proposals
  - anoncake 7y ago
    Reading this made me think "The Register?" scrolls up The Register.
tssva 7y ago
> The proposal is consistent with Apple's attempt to occupy the moral high-ground of technology by championing privacy at the expense of the surveillance capitalism embodied by Google and Facebook. Shouldn't this be "The proposal is consistent with Apple's attempt to disrupt the largest revenue stream of their main competitor Google." If Apple actually cared about "the moral high-ground of technology' they wouldn't be so desperately fighting right to repair laws.
- taneq 7y ago
  To-may-to, to-mar-to.
- oarsinsync 7y ago
  I was wondering why you were being downvoted, as the point seems pretty valid. Privacy is as big a concern to the vast majority as right-to-repair is. I suspect that right-to-repair occupies an even smaller mindshare than privacy does these days too. Eitherway, it's all sub-par, and that's a real shame. Then I read the article, and realised the paragraph you're partially quoting is saying the same thing: > The proposal is consistent with Apple's attempt to occupy the moral high-ground of technology by championing privacy at the expense of the surveillance capitalism embodied by Google and Facebook. Note this is for Western iThing users only, if you're a Chinese customer privacy is just a distant dream. Unfortunately, while The Register's entire article is written in that tone, when a comment does it, it doesn't translate quite as well.
- simonh 7y ago
  If right to repair really was just about user rights that would be fair enough, but it’s also about requiring bigger, heavier, more fragile and more expensive products. As a user I’m all for user rights, but I personally doubt the advantages of the government telling companies how to design their products beyond things like health, safety and environmental protection. I don’t think users should have to worry about products being safe for example, but I think the trade off between repairability versus other desirable attributes is something best left to user choice.
  - 7y ago
founderling 7y ago
Traditionally, ad click attribution has been done through the use of cookies and so-called “tracking pixels.” No. Clicks are not tracked via cookies but via urls. The search engine in their example would send the user to someshop.com/someproduct?clickid=7e82jv927x748342 They say nothing about how they want to prevent this and other tracking mechanisms. Yet, they propose an overly complex system to send even more data to advertisers. Also they do not say anything about the ip that their additional ping will send out. I definitely do not want my browser to communicate with an advertiser days later and without my consent. Also, click tracking is not even a big problem in the first place. Tracking you wherever you go is. Even if you do not click on any ads.
- om2 7y ago
  We (WebKit team) know of major ad platforms that use cookies as their preferred mechanism to track clicks, in part so that a tracking pixel can work without requiring the merchant landing and conversion pages to include script. We are also aware of tracking via link decoration (not just for ads) and our first steps at defending against it are described here: https://webkit.org/blog/8828/intelligent-tracking-prevention-2-2/ https://webkit.org/blog/8828/intelligent-tracking-prevention...
libertine 7y ago
So... the contextual advertising is not sticking uh?
makomk 7y ago
"Once the browser has matched a conversion against a stored ad click, it sets a timer, randomized between 24 and 48 hours. When that timer fires, the browser makes an ephemeral, stateless POST request to the same well-known location." Alternatively, once the ad fraudsters have decided to scam an advertiser using this system, their ad fraud programs make a series of POST requests to the same well-known location pretending to be copies of Safari that have seen conversions from this ad campaign. The only way to detect this fraud would be to match up actual orders with claimed conversions from a particular site, completely bypassing the intended privacy protections. (Which probably wouldn't be that hard in some circumstances, but fraudsters would of course get to target the scenarios where matching up orders is hardest.) Like, as far as I can tell, this completely fails to provide attribution data that advertisers can trust at all because it relies totally on the honesty of software installed on end-user devices. It will also fail to provide the stated privacy level in many situations, such as when the user keeps the same unique-ish IP address for several days. The only thing it seems to achieve is good PR for Apple, who will get a bunch of stories from credulous reporters about how they're trying to improve user privacy and the evil adtech industry is thwarting them because it wants to know everything about you.
- stingraycharles 7y ago
  Yeah I don't think this whole proposal works at all, and shows these people have very little experience with the actual adtech industry. To me it seems like they could only get a proposal this wrong if they did not even consult a single industry expert.
- lapnitnelav 7y ago
  Yeah let's not even mention the super low count of campaigns and conversions an advertiser can have at any given time. Some of the advertisers we work with have pretty small budget, i.e. less than 50k/month and even them will probably max out the 64 campaigns. I'm all for having a privacy minded advertising industry but this proposal misses the mark big time.
  - makecheck 7y ago
    The way I see it, the ad industry “survived” just fine when relying only on print ads and other mechanisms that had no concept of automatically-logged impressions. Ad companies do not have an inherent right to abuse tech just because tech exists.
Someone 7y ago
"Google can scan users' Gmails to see what items they bought," he said. "That is why Amazon removed the list of products from order confirmation emails and require the user to click and login to Amazon to see the order details. Does Google do that? If so, only for a limited set of email senders, or globally? Would they, for example, read email sent between doctors and patients?
- msbarnett 7y ago
  Google absolutely does that: https://myaccount.google.com/purchases https://myaccount.google.com/purchases They scan your email and extract what you purchased, when, for how much, and when it was delivered. It appears to be as close to universal as they can get it — they’ve extracted info from some pretty niche retailers emails on my account. There’s no publicly viewable equivalent for scanned health info but, internally, who knows. It wouldn’t be at all surprising if some of that data went into a training set for ad targeting, at the very least.
  - jobigoud 7y ago
    Wow, that's eerie. And I thought I had everything disabled. Can this be disabled?
    - leokennis 7y ago
      Yes. You can disable it by switching to Fastmail or a similar email provider. (Sorry.)
      - Someone 7y ago
        And never send email to gmail accounts, and make sure nobody ever sends you mail from gmail accounts, I presume? That’s quite unworkable.
        leokennis 7y ago
        Of course you’re right; perfect privacy is impossible. But at least we can pick the low hanging fruit and not give our own info to Google voluntarily.
IloveHN84 7y ago
Simple reminder that this can't prevent browser fingerprinting, making this technique useless
- om2 7y ago
  We've done things to make it harder fingerprint Safari users. It might still work sometimes but will be less reliable.

125 points by fauigerzigerk 7y ago

Google has a million ways of tracking us anyway, this is going to be just a minor problem for them they will solve sooner or later.

bdibs 7y ago
Does that mean people shouldn’t try to stop it?
- dvfjsdhgfv 7y ago
  No, I think we should try anyway, but it will be extremely difficult to do and a lot more effort should be put in educating people about what price they pay for "free" product handing over their and others' data into the hands of a single entity.
StreamBright 7y ago
Actually there are ways to stop Google like pi-hole for example
- billyhastings 7y ago
  Facebook ads regularly bypass pi-hole (at least in my experience). However, I disagree entirely with the assumption of "just because they find way we shouldn't try". With that attitude, your ancestors would likely be either Japanese or German.
- sixothree 7y ago
  I wonder how many people they have working to circumvent pi-hole specifically.
  - adrianN 7y ago
    As roughly zero people who would click an ad use a pihole, I'd guess roughly zero engineers are working on the problem.
  - StreamBright 7y ago
    How would you go about "circumventing" pi-hole? I do not allow DNS requests going towards Google's DNS servers by firewall. The only option would be to use a DoH but it is also blocked on standard ports. The only option is to use a non-Google associated IP with a non-standard port which is not worth it at the scale of Google.

AJRF 7y ago

Forgive the pessimism but isn't this article rewritten every year? It's always along the lines of "ad tech company begs apple not to enable X". Doesn't really seem to be that effective as there is always something to replace the newly blocked method each time.

paranoidrobot 7y ago
That would probably be why the subheading is: > Safari tech ready to be ignored by online ad giants like all other privacy proposals
- anoncake 7y ago
  Reading this made me think "The Register?" scrolls up The Register.

tssva 7y ago

> The proposal is consistent with Apple's attempt to occupy the moral high-ground of technology by championing privacy at the expense of the surveillance capitalism embodied by Google and Facebook. Shouldn't this be "The proposal is consistent with Apple's attempt to disrupt the largest revenue stream of their main competitor Google." If Apple actually cared about "the moral high-ground of technology' they wouldn't be so desperately fighting right to repair laws.

taneq 7y ago
To-may-to, to-mar-to.
oarsinsync 7y ago
I was wondering why you were being downvoted, as the point seems pretty valid. Privacy is as big a concern to the vast majority as right-to-repair is. I suspect that right-to-repair occupies an even smaller mindshare than privacy does these days too. Eitherway, it's all sub-par, and that's a real shame. Then I read the article, and realised the paragraph you're partially quoting is saying the same thing: > The proposal is consistent with Apple's attempt to occupy the moral high-ground of technology by championing privacy at the expense of the surveillance capitalism embodied by Google and Facebook. Note this is for Western iThing users only, if you're a Chinese customer privacy is just a distant dream. Unfortunately, while The Register's entire article is written in that tone, when a comment does it, it doesn't translate quite as well.
simonh 7y ago
If right to repair really was just about user rights that would be fair enough, but it’s also about requiring bigger, heavier, more fragile and more expensive products. As a user I’m all for user rights, but I personally doubt the advantages of the government telling companies how to design their products beyond things like health, safety and environmental protection. I don’t think users should have to worry about products being safe for example, but I think the trade off between repairability versus other desirable attributes is something best left to user choice.
- 7y ago

founderling 7y ago

Traditionally, ad click attribution has been done through the use of cookies and so-called “tracking pixels.” No. Clicks are not tracked via cookies but via urls. The search engine in their example would send the user to someshop.com/someproduct?clickid=7e82jv927x748342 They say nothing about how they want to prevent this and other tracking mechanisms. Yet, they propose an overly complex system to send even more data to advertisers. Also they do not say anything about the ip that their additional ping will send out. I definitely do not want my browser to communicate with an advertiser days later and without my consent. Also, click tracking is not even a big problem in the first place. Tracking you wherever you go is. Even if you do not click on any ads.

om2 7y ago
We (WebKit team) know of major ad platforms that use cookies as their preferred mechanism to track clicks, in part so that a tracking pixel can work without requiring the merchant landing and conversion pages to include script. We are also aware of tracking via link decoration (not just for ads) and our first steps at defending against it are described here: https://webkit.org/blog/8828/intelligent-tracking-prevention-2-2/ https://webkit.org/blog/8828/intelligent-tracking-prevention...

libertine 7y ago

So... the contextual advertising is not sticking uh?

makomk 7y ago

"Once the browser has matched a conversion against a stored ad click, it sets a timer, randomized between 24 and 48 hours. When that timer fires, the browser makes an ephemeral, stateless POST request to the same well-known location." Alternatively, once the ad fraudsters have decided to scam an advertiser using this system, their ad fraud programs make a series of POST requests to the same well-known location pretending to be copies of Safari that have seen conversions from this ad campaign. The only way to detect this fraud would be to match up actual orders with claimed conversions from a particular site, completely bypassing the intended privacy protections. (Which probably wouldn't be that hard in some circumstances, but fraudsters would of course get to target the scenarios where matching up orders is hardest.) Like, as far as I can tell, this completely fails to provide attribution data that advertisers can trust at all because it relies totally on the honesty of software installed on end-user devices. It will also fail to provide the stated privacy level in many situations, such as when the user keeps the same unique-ish IP address for several days. The only thing it seems to achieve is good PR for Apple, who will get a bunch of stories from credulous reporters about how they're trying to improve user privacy and the evil adtech industry is thwarting them because it wants to know everything about you.

stingraycharles 7y ago
Yeah I don't think this whole proposal works at all, and shows these people have very little experience with the actual adtech industry. To me it seems like they could only get a proposal this wrong if they did not even consult a single industry expert.
lapnitnelav 7y ago
Yeah let's not even mention the super low count of campaigns and conversions an advertiser can have at any given time. Some of the advertisers we work with have pretty small budget, i.e. less than 50k/month and even them will probably max out the 64 campaigns. I'm all for having a privacy minded advertising industry but this proposal misses the mark big time.
- makecheck 7y ago
  The way I see it, the ad industry “survived” just fine when relying only on print ads and other mechanisms that had no concept of automatically-logged impressions. Ad companies do not have an inherent right to abuse tech just because tech exists.

Someone 7y ago

"Google can scan users' Gmails to see what items they bought," he said. "That is why Amazon removed the list of products from order confirmation emails and require the user to click and login to Amazon to see the order details. Does Google do that? If so, only for a limited set of email senders, or globally? Would they, for example, read email sent between doctors and patients?

msbarnett 7y ago
Google absolutely does that: https://myaccount.google.com/purchases https://myaccount.google.com/purchases They scan your email and extract what you purchased, when, for how much, and when it was delivered. It appears to be as close to universal as they can get it — they’ve extracted info from some pretty niche retailers emails on my account. There’s no publicly viewable equivalent for scanned health info but, internally, who knows. It wouldn’t be at all surprising if some of that data went into a training set for ad targeting, at the very least.
- jobigoud 7y ago
  Wow, that's eerie. And I thought I had everything disabled. Can this be disabled?
  - leokennis 7y ago
    Yes. You can disable it by switching to Fastmail or a similar email provider. (Sorry.)
    - Someone 7y ago
      And never send email to gmail accounts, and make sure nobody ever sends you mail from gmail accounts, I presume? That’s quite unworkable.
      - leokennis 7y ago
        Of course you’re right; perfect privacy is impossible. But at least we can pick the low hanging fruit and not give our own info to Google voluntarily.

IloveHN84 7y ago

Simple reminder that this can't prevent browser fingerprinting, making this technique useless

om2 7y ago
We've done things to make it harder fingerprint Safari users. It might still work sometimes but will be less reliable.