Google's Federated Learning of Cohorts (FLoC) was proposed as a privacy-preserving alternative to third-party cookies. Instead of tracking individual users, FLoC would group users into cohorts based on browsing behavior and allow advertisers to target these groups. The proposal was met with widespread rejection.

Why FLoC Was Rejected

Privacy advocates pointed out that cohort IDs could still be used for fingerprinting and cross-site tracking. Interest-based cohorts could reveal sensitive information about users' health conditions, political beliefs, or sexual orientation. Browsers other than Chrome refused to implement FLoC, and privacy organizations, regulators, and even other advertising companies opposed it.

The Replacement: Topics API

After FLoC's failure, Google proposed the Topics API as a replacement. Instead of cohorts, Topics assigns users to broad interest categories based on recent browsing. While less granular than FLoC, Topics faced similar criticisms about its ability to leak sensitive information and its potential for abuse.

The Outcome

Both Privacy Sandbox proposals failed to gain acceptance, and Google ultimately abandoned its plan to deprecate third-party cookies entirely. Chrome remains the only major browser that accepts third-party cookies by default, while Safari, Firefox, and Brave have blocked them for years.

The Lesson

The FLoC saga demonstrates the fundamental difficulty of replacing invasive tracking with slightly-less-invasive tracking. The advertising industry's attempts to find privacy-compatible tracking mechanisms have repeatedly failed because the underlying goal -- monitoring user behavior for ad targeting -- conflicts with genuine privacy.