RC RANDOM CHAOS

Firefox 143: Fingerprinting Identifiable Users Cut from 75% to 46%

· via Hacker News

Original source

A comprehensive study of how fingerprintable users are

Hacker News →

Mozilla researchers analyzed browser fingerprinting across almost 25 million Firefox users to understand its real-world impact. The study identified that user uniqueness results from a small set of high-impact characteristics rather than accumulation of many independent measurements. The fingerprinter Fingerprint Pro identifies 80% of its targeted unique users with just 8 metrics from its total collection of 50+, while Signifyd achieves 80% identification using 9 of its 80+ metrics.

Mozilla developed protections shipped in Firefox 143 that reduced uniquely identifiable users from 75% to 46% while preserving web compatibility. Building effective protections requires both standardization and randomization, though neither approach alone provides complete privacy. Randomization is potent when functional but can be detected and bypassed. Protection effectiveness varies significantly across operating systems—screen resolution clamping removes 25.3 percentage points on macOS but only 0-3 points on others.

The measurement process used privacy-preserving technologies like OHTTP to prevent individual user identification. The analysis measured cohort structure using the Simpson Index and found that Shannon entropy alone does not reliably indicate fingerprinting effectiveness. This approach revealed protections reducing uniqueness equally can have vastly different impacts on user grouping. Root cause analysis of metric variations informed improvements to image rendering, Web Audio, and mathematical operations on Android.

Read the full article

Continue reading at Hacker News →

This is an AI-generated summary. Read the original for the full story.