Browser Fingerprinting

Browser fingerprinting is the technique of identifying a browser by combining many small, individually unremarkable signals into a unique identifier. The most common signals include canvas rendering output, WebGL parameters, installed fonts, screen resolution, timezone, language settings, navigator properties, audio context characteristics, and the precise list of supported plugins and codecs. Combined, these produce a fingerprint stable enough to track a user across sessions even without cookies, and distinctive enough to flag automated traffic that does not match normal human distributions.

Anti-bot systems use fingerprinting both for tracking (the same fingerprint repeatedly hitting a sensitive endpoint is suspicious) and for detection (a headless browser fingerprint differs from a real user's in many subtle ways, even after spoofing the user agent). Tools like FingerprintJS make commercial-grade fingerprinting available to any site; sophisticated targets like banking and travel sites build their own.

For AI builders running web automation, fingerprint hardening is an active area of work. Stealth plugins for Puppeteer and Playwright patch known automation tells; managed browser infrastructure providers (Browserbase, Steel.dev, Anchor Browser) ship pre-hardened browser images that are harder to fingerprint as automation. The practical advice is the same as with anti-bot detection: pay a vendor whose business is keeping the fingerprint surface clean.