🚩 Red Flag

🧠 Inspiration

Every time we click “I Agree,” we’re entering a contract we’ve never read. Privacy policies are designed to exhaust us — filled with jargon, loopholes, and 10,000-word sentences that no sane human enjoys.

We realized this is a red flag moment — we’re handing over data, location, even behavior, without informed consent.

So we built Red Flag — a Chrome extension that reads the privacy policy so you don’t have to. It translates the fine print into plain English, gives a risk score, and automatically displays it right above the signup form.

Our goal? To make online privacy transparent, fast, and understandable — for everyone.

⚙️ How We Built It

Red Flag is a blend of natural language processing, frontend automation, and browser integration magic.

Here’s the flow:

Policy Detection: When a user lands on a signup page, the extension scans for links that match patterns like “/privacy” or “/policy”.
Text Extraction: The policy text is fetched, cleaned, and stripped of HTML noise.
NLP Analysis: We used a transformer-based model fine-tuned to detect key “red flag” phrases — like

“share with partners”
“retain data indefinitely”
“collect personal identifiers” Each phrase is weighted and scored.

The overall risk score is calculated as: $$ \text{Risk Score} = \frac{\sum_i w_i \cdot f_i}{\sum_i w_i} $$ where ( f_i ) is the frequency of a risk phrase and ( w_i ) its severity weight.

Summarization: We generate a short, human-readable summary using a text simplification pipeline.
Frontend Injection: Using Chrome’s content scripts, we inject a sleek widget right above the signup form showing:

⚠️ Risk Score (0–100)
🧾 Simplified Summary
🔍 “View Full Breakdown” button

Future Vision: We’re expanding Red Flag to analyze your digital footprint — mapping every app and website you’ve given permissions to, with a privacy “health score.”

🧩 What We Learned

Privacy policies are sneakier than we thought. Even legitimate companies use vague phrasing to hide data collection.
NLP can be ethical. Building tech that protects users instead of exploiting them was incredibly empowering.
UX matters. People won’t use privacy tools unless they’re simple, fast, and invisible until needed.
Automation is key. The “auto-insert” feature made it feel like magic — privacy warnings appear before you make a mistake.

🚧 Challenges We Faced

Parsing nightmares: Privacy policies vary wildly in structure, formatting, and accessibility.
Balancing accuracy with simplicity: Too technical = users tune out; too simple = loses meaning.
CORS & browser restrictions: Fetching external content from within Chrome required clever workarounds.
Scoring logic: Quantifying privacy risk isn’t objective — we had to design a scoring model that felt fair and intuitive.
Time pressure: We built all of this within hackathon constraints — from NLP pipeline to Chrome extension UI.

🌍 The Vision

We don’t want privacy awareness to be optional — we want it to be default.

Red Flag is more than an extension — it’s a movement towards digital transparency. Because in 2025, the biggest “red flag” online… is still that tiny “I Agree” button. 🚩