GlassBox: Unmasking Browser Fingerprinting & User Identifiability

GlassBox is a client-side browser fingerprinting benchmark designed to reveal the exact data points websites use to track you. Instead of these signals remaining ~~hidden~~ surfaced, GlassBox runs every possible measurement live and displays the results directly to the user.

🔍 How Tracking Works: The Hierarchy of Identification

Tracking scripts and anti-fraud systems don't just look at one thing; they use a layered approach to determine who you are. GlassBox categorizes these into four distinct tiers:

Tier	Focus	Scope	Persistence
1	Hardware Environment	Cross-Engine	Links you across different browsers
2	Engine $\times$ Hardware	Engine Family	Links you within the same engine (e.g., Chromium)
3	Browser Build	Versioning	Identifies the software, not the person
4	Session Data	Temporary	Resets frequently (IP, Theme, Timing)

The Identification Logic

The probability of being uniquely identified can be expressed as a function of the entropy provided by these tiers: $\text{Identifiability} \approx \sum (\text{Hardware} + \text{Engine} + \text{Build} + \text{Session})$

🛠️ Features & Functionality

GlassBox doesn't just measure; it allows you to analyze and mitigate your digital footprint.

Identifiability Scoring: The tool calculates whether you appear as the same machine, the same engine family, or merely the same browser version.
Comparative Analysis: You can determine if you are "re-linkable" by importing data.
- Action: Copy a JSON report from one session $\rightarrow$ Paste it into another.
Educational Resources: Includes a comprehensive how-to guide on reducing your fingerprint.

Available Toolset:

🔒 Privacy & Data Handling

The vast majority of the audio-fp probes and hardware checks are executed entirely within your browser. There are no beacons and no analytics sent to a remote server.

The Single Exception: To provide IP intelligence, GlassBox must communicate with external public APIs:

ipwho.is
ipapi.is

These are used specifically to resolve your IP address, identify your network provider, and detect if you are using a VPN.

💻 Technical Output Example

When you export your data, GlassBox generates a structured report similar to this:

{
  "fingerprint": {
    "hardware": "Generic-X86_64",
    "engine": "Blink",
    "canvas": "a1b2c3d4e5f6",
    "audio": "f9e8d7c6b5a4",
    "session": {
      "theme": "dark",
      "ip_resolved": "1.2.3.4"
    }
  },
  "identifiability_score": "High"
}

For more details, users can Expand notes, Re-run probes, or utilize the Copy JSON report feature to audit their privacy across different environments.