3 ms·

Yes, if you assume the people who designed this API were idiots, and everyone who reviewed the proposals were idiots, that could be possible outcome. (Not a lik

by jsnell 3y ago

Yes, if you assume the people who designed this API were idiots, and everyone who reviewed the proposals were idiots, that could be possible outcome. (Not a likely outcome, mind you, because it'd be an incredibly low-value and low-yield attack). But why just assume that, rather than read the actual proposal? In reality, the topics come from a predefined taxonomy[0] that certainly won't include topics like "bondage", nor (by policy) other kinds of sensitive PII. Additionally the system inserts noise into the system (5% of the time the topic will be a random choice rather than one of your actual inferred topics of interest), so that you can't actually infer the user's real interests from a single ad click like that. [0] https://github.com/patcg-individual-drafts/topics/blob/main/taxonomy_v1.md https://github.com/patcg-individual-drafts/topics/blob/main/... [1] https://developer.chrome.com/docs/privacy-sandbox/topics/overview/#how-it-works https://developer.chrome.com/docs/privacy-sandbox/topics/ove...

tikhonj 3y ago
You don't have to assume they're idiots, you just have to assume they aren't entirely serious about privacy.
dspillett 3y ago
*> by policy And we all know that policies, like end user agreements, never change. They are especially never changed unilaterally by commercial entities for which the change could be beneficial to their business plans (directly, or by virtual of a lobbying party negotiating an exchange of conveniences).
amelius 3y ago
Even if what you say is true, they still leak my data! Also, there's a lot you can do with innocent looking data, simply through correlations. Just like people like to generalize from skin color (I mean, what is more innocent than skin color?) I'm pretty sure that you can even infer skin color from things like "likes R&B music" and a few similar datapoints. Of course, you don't get exact information but statistics can be enough to hurt someone. That's also why your noise argument doesn't make much sense.
josefx 3y ago
> [0] https://github.com/patcg-individual-drafts/topics/blob/main/ https://github.com/patcg-individual-drafts/topics/blob/main/... Nice v1 at the end. We can assume that this list is final and will not be changed? > nor (by policy) other kinds of sensitive PII. Yeah, it just exposes interest in family planing, loans, ..., which we do know have absolutely no potential for abuse. Or given the attempts to outlaw drag there is probably no potential way to use interest in Nail Care or Makeup in a negative fashion, right? > Yes, if you assume the people who designed this API were idiots I assume they are getting paid well to play the role.
- stavros 3y ago
  I don't disagree with your argument, but I think you're making it in a way that's tautological, and a bit disingenuous. You're saying "this API has privacy concerns" and then point to some privacy concerns. However, you aren't taking into account that it's not this specific API that has privacy concerns. Any advertising API would have privacy concerns, because it's explicit goal is to be able to target users for ads. Pointing to some stuff and saying "this API that's made to target users allows user targeting" isn't bringing anything new, really.
scarface_74 3y ago
I’ll just choose to have a browser that doesn’t have this “feature”…

by jsnell 3y ago

tikhonj 3y ago

You don't have to assume they're idiots, you just have to assume they aren't entirely serious about privacy.

dspillett 3y ago

*> by policy And we all know that policies, like end user agreements, never change. They are especially never changed unilaterally by commercial entities for which the change could be beneficial to their business plans (directly, or by virtual of a lobbying party negotiating an exchange of conveniences).

amelius 3y ago

Even if what you say is true, they still leak my data! Also, there's a lot you can do with innocent looking data, simply through correlations. Just like people like to generalize from skin color (I mean, what is more innocent than skin color?) I'm pretty sure that you can even infer skin color from things like "likes R&B music" and a few similar datapoints. Of course, you don't get exact information but statistics can be enough to hurt someone. That's also why your noise argument doesn't make much sense.

josefx 3y ago

> [0] https://github.com/patcg-individual-drafts/topics/blob/main/ https://github.com/patcg-individual-drafts/topics/blob/main/... Nice v1 at the end. We can assume that this list is final and will not be changed? > nor (by policy) other kinds of sensitive PII. Yeah, it just exposes interest in family planing, loans, ..., which we do know have absolutely no potential for abuse. Or given the attempts to outlaw drag there is probably no potential way to use interest in Nail Care or Makeup in a negative fashion, right? > Yes, if you assume the people who designed this API were idiots I assume they are getting paid well to play the role.

stavros 3y ago
I don't disagree with your argument, but I think you're making it in a way that's tautological, and a bit disingenuous. You're saying "this API has privacy concerns" and then point to some privacy concerns. However, you aren't taking into account that it's not this specific API that has privacy concerns. Any advertising API would have privacy concerns, because it's explicit goal is to be able to target users for ads. Pointing to some stuff and saying "this API that's made to target users allows user targeting" isn't bringing anything new, really.

scarface_74 3y ago

I’ll just choose to have a browser that doesn’t have this “feature”…