Thread by @WolfieChristl

Do I get this right that the current state of Google et al's TURTLEDOVE/FLEDGE proposal would lead to browsers putting users into myriads of 'interest groups', still allowing advertisers, publishers and third-party adtech companies to almost act 1:1 on specific user behavior?

...as long as 100 users behave similarly, which is not a really tough restriction. Almost no remarketing/etc campaign addresses less than 100 potential targets.

Fledge:
github.com/WICG/turtledov…

Turtledove:
github.com/WICG/turtledov…

I didn't have the time to follow the developments :/

Generally, I feel like Google and a few other players are working on this fundamental effort to preserve/perpetuate a web economy based on behavioral advertising without any relevant participation of non-industry stakeholders.

My understanding is that Chrome would:

1) put users into general clusters with similar 'interests' based on federated learning:
github.com/WICG/floc

2) replace everything advertisers, marketers and adtech firms now use 'audiences' for with the stuff I was discussing above

3) preserve different ways to measure whether people are acting how marketers want them to act and who gets paid for it. No behavioral advertising without 'conversion measurement' and 'attribution':
github.com/WICG/conversio…
github.com/csharrison/agg…

4) implement a distributed 'privacy-preserving' fraud prevention panopticon to make sure marketers only spend on commercially valuable users:
github.com/WICG/trust-tok…

(yes detecting invalid traffic is necessary on the web, but reducing ad fraud should not be in the center of it)

5) make everyone login into the browser

I welcome the (possible) death of individual-level digital profiling based on third-party cookies and ubiquitous personal data leakage across companies and contexts.

However.

First, as this will increase platform power even more, policymakers must ban Google et al from further exploiting data only they have access in several ways.

Second, I'm guess Google will a) keep tracking alive as long as possible, also on mobile b) further weaken its proposals.

c) based on my understanding, it will in many ways perpetuate how personalization for profit affects individuals, groups and society, from Kafkaesque experiences to systemic discrimination to behavioral feedback loops to manipulation to perpetuating a broken web economy at large.

Anyway, the large-scale misuse of individual-level data must vanish from the ecosystem, so I welcome approaches that go that way. Don't listen to ppl who say this is already done. It's not.

On top of that, there are additional challenges.

In any case, Google must be broken up.

To return to my initial question, if the below 'use case' would be possible, how is this not just personal data processing as defined in the GDPR, regardless of whether it is stored in the client and regardless of whether third parties get granular data?
github.com/WICG/turtledov…