Inspiration

I realized that every time we browse the web, we act as unpaid data laborers for AI companies. Our personal identities are harvested to train models we do not own or control. I wanted to build a tool that moves beyond passive blocking and allows users to actively protect their digital sovereignty through adversarial noise.

What it does

Sentinel OS is a privacy layer that poisons the data packets unauthorized AI scrapers try to collect. It detects fingerprinting attempts and injects randomized telemetry, making the user's digital footprint useless for AI training sets.

How we built it

Frontend: Built a premium, minimal dashboard using React. Backend: Integrated Firebase for secure user authentication and tracking real-time protection statistics. Engine: Developed a Manifest V3 Chrome extension using content scripts to scramble Canvas and WebGL fingerprints.

Challenges we ran into

The biggest challenge was balancing "data poisoning" with browser performance. Injecting noise into every outgoing request can cause lag, so I optimized the scripts to only fire when high-confidence scraper behavior is detected.

Accomplishments that we're proud of

I successfully developed a working prototype that triggers "Automated behavior detected" on professional-grade fingerprint scanners. This proves that we can effectively pollute the training data of predatory AI models.

What we learned

Building Sentinel OS taught me the technical depth of Adversarial Data Privacy. I learned how modern AI scrapers use high-fidelity fingerprinting to bypass standard VPNs and ad-blockers. On the development side, I mastered Chrome’s Manifest V3 API and learned how to manage real-time data states between a React frontend and a background content script using Firebase. Most importantly, I learned that the most effective way to protect digital identity in 2026 isn't to hide data, but to make it useless for unauthorized training.

What's next for Sentinel OS

I plan to scale Sentinel OS from a browser extension into a standalone, privacy-first browser kernel. My goal is to protect the next billion users from digital identity harvesting.

Built With

Share this project:

Updates

Submission history