Third-party cookies are dying. Google finally deprecated them in Chrome in late 2025, joining Firefox and Safari which blocked them years earlier. The tracking industry didn’t panic — they’d already shifted to fingerprinting.

Browser fingerprinting identifies you by collecting dozens of data points about your browser and device: screen resolution, installed fonts, WebGL renderer, audio context properties, timezone, language, platform, and many more. Combined, these create a fingerprint that’s often unique enough to track you across websites without storing anything on your device.

We wanted to know: in 2026, with modern browsers and anti-fingerprinting tools, how trackable are you actually?

Our Testing Setup

We tested fingerprint uniqueness using AmIUnique and the EFF’s Cover Your Tracks across multiple configurations:

• Chrome (default settings)
• Firefox (default settings)
• Firefox (with resistFingerprinting enabled)
• Brave (default shields)
• Tor Browser
• Safari (default settings)
• Chrome with Canvas Blocker extension
• Mullvad Browser

Each test was run from two different machines (Windows 11 desktop, MacBook Pro) on a residential connection.

Default Browser Results

Chrome (default): Fully unique fingerprint on both machines. Canvas, WebGL, and AudioContext all returned distinct values. Screen resolution, timezone, installed fonts, and platform created an easily identifiable profile. Among the AmIUnique database, our Chrome fingerprint matched 0 other users. Total uniqueness.

Firefox (default): Also unique, but slightly less so. Firefox blocks some fingerprinting vectors by default now — it restricts font enumeration and resists some canvas fingerprinting. But the overall fingerprint was still unique in the database.

Safari (default): Apple has implemented Intelligent Tracking Prevention which includes some fingerprint randomisation. Our Safari fingerprint was less unique than Chrome or Firefox, but still identifiable among a pool of approximately 200 similar configurations. Not anonymous by any means.

Brave (default shields): Significantly better. Brave randomises canvas and WebGL fingerprints by default. Our fingerprint wasn’t unique — it matched a small pool of other Brave users. Brave’s approach of adding noise to fingerprinting vectors rather than blocking them outright works well.

Anti-Fingerprinting Configurations

Firefox with resistFingerprinting (privacy.resistFingerprinting = true): Major improvement. This setting spoofs timezone to UTC, reports a standardised screen size, blocks font enumeration, and returns uniform canvas/WebGL values. Our fingerprint matched a pool of thousands of other resistFingerprinting users. The trade-off: some websites break. Date pickers show wrong timezone. Layout can look odd due to reported screen dimensions.

Tor Browser: As expected, the strongest result. Tor Browser is specifically designed so all users look identical. Our fingerprint was indistinguishable from any other Tor Browser user at the same window size. The trade-off: speed is poor, many sites block Tor exit nodes, and the browsing experience is noticeably degraded.

Mullvad Browser: Built on the same principles as Tor Browser but without the Tor network. Excellent fingerprint resistance — our configuration matched a large pool. Practical for daily use with a VPN. This was the most usable option with strong fingerprint resistance.

The Most Revealing Fingerprint Vectors

Not all fingerprinting data points are equally identifying. Based on our testing, these contributed most to uniqueness:

Canvas fingerprinting: Your GPU renders text and shapes slightly differently from every other GPU. This alone can narrow identification to a small pool. Brave’s noise injection and Firefox’s resistFingerprinting both address this.

WebGL renderer string: Reports your exact GPU model. “ANGLE (Apple, Apple M2 Pro, OpenGL 4.1)” is quite specific. Combined with other data, very identifying.

Installed fonts: Despite restrictions, font enumeration still partially works. The set of fonts on your system varies enough to be identifying, especially on Windows where users install custom fonts.

Screen resolution and colour depth: 1920x1080 is common enough to provide some anonymity. Unusual resolutions (ultrawide monitors, scaled HiDPI) are more identifying. The combination with device pixel ratio narrows things further.

AudioContext fingerprint: Your audio stack processes signals in hardware-specific ways. This fingerprint is stable across sessions and hard to spoof without breaking web audio.

What Actually Works

Based on our testing, here’s what reduces fingerprint trackability in practice:

Tier 1 (most protection): Tor Browser or Mullvad Browser. Near-complete fingerprint uniformity. Tor Browser is more anonymous but less usable. Mullvad Browser is the practical choice paired with Mullvad VPN.

Tier 2 (strong protection): Brave with default shields, or Firefox with resistFingerprinting enabled. Both significantly reduce uniqueness. Brave has fewer usability trade-offs.

Tier 3 (moderate protection): Firefox with Enhanced Tracking Protection set to Strict mode. Blocks known fingerprinting scripts but doesn’t address all vectors.

Tier 4 (minimal protection): Chrome with extensions. Canvas blockers help with one vector but leave dozens of others exposed. The sense of protection exceeds the reality.

The Extension Paradox

Here’s something counterintuitive: installing privacy extensions can make you more fingerprintable. If you install Canvas Blocker, uBlock Origin, Privacy Badger, and Decentraleyes, the specific combination and configuration of those extensions becomes part of your fingerprint.

You’ve gone from “one of millions of default Chrome users” to “one of hundreds of users with this exact extension stack.” The extensions block some tracking vectors while creating new identifying characteristics.

The lesson: fewer, well-chosen tools beat a stack of extensions. A single well-designed browser (Brave, Mullvad) does more than Chrome with five privacy add-ons.

What This Means Practically

Complete fingerprint anonymity while maintaining normal web browsing isn’t achievable in 2026. The web platform exposes too many identifying data points, and websites increasingly depend on the APIs that enable fingerprinting.

But you can move from “uniquely identifiable” to “lost in a crowd.” That shift — from a pool of 1 to a pool of thousands — is meaningful. It forces trackers from deterministic identification (they know it’s you) to probabilistic identification (they think it might be you).

For most people, switching to Brave or enabling Firefox’s stricter protections provides substantial improvement with minimal disruption. For those with higher privacy requirements, Mullvad Browser offers a serious tool without Tor’s speed penalties.

The tracking industry is sophisticated and well-resourced. But you don’t need to beat every tracker. You need to make tracking expensive and unreliable enough that you’re not worth the effort. That’s achievable with the right browser choice.