A developer has released Glassbox, a free tool that runs entirely in a user's browser to reveal how easily companies can fingerprint and track individuals online, without sending any data to external servers. Built by security engineer David Dale with assistance from Claude Code, the tool matters because it exposes the many subtle techniques, from audio analysis to canvas rendering, that trackers use to identify users even without cookies, and it offers practical advice on reducing that exposure.

Glassbox checks more than 30 signals, including canvas, WebGL, font libraries, WASM features and cross-site login states, including a newly noted technique from AliExpress that uses silent audio waveforms to identify browsers. Testing by The Register found identifiability estimates ranging from 99 percent in Chrome down to 56 percent in Tor Browser, with Firefox and Tor the only browsers that avoided being labelled "effectively unique". Dale stressed the figures are estimates rather than precise measurements, since the tool has no live population data to compare against, unlike rivals AmIUnique and EFF's Cover Your Tracks. His key advice is that blending into a large crowd of similar browser configurations, rather than using a heavily customised "hardened" setup, offers the best protection, alongside using a VPN or Tor and blocking WebRTC leaks.