AI Genesis Challenge • Level 1: Spot the Glitch

ClauseGuard — Stop Signing Away Your Money and Privacy in 0.2 Seconds

Subheadline: Everyone clicks "I Agree" without reading the fine print. ClauseGuard reads 15-page policies in seconds and spotlights the hidden traps before you commit.


1. The Glitch (The Problem)

Every time you download a free app, start a "free" trial, or create an account, you are handed a wall of dense legal jargon designed to be skipped. Because nobody has 30 minutes to decode fine print before signing up, regular people repeatedly get burned by the exact same hidden traps:

The Core Insight: People don't ignore Terms of Service because they don't care—they ignore them because legal documents aren't written for humans.

2. The Solution (How It Works)

Instead of forcing users to become lawyers, ClauseGuard acts like a red-pen highlighter for fine print:

  1. Drop a Link or Paste Text: Before clicking "Accept" on any website or app, paste the Terms of Service link (or copy the text).
  2. Instant Trap Score (1–10): Get an immediate, color-coded safety rating so you know in two seconds whether a policy is fair or predatory.
  3. Plain-English Callouts + Exact "Receipts": Every red flag—like Auto-Renewal Trap or Data Selling—is translated into a single plain-English sentence, paired with the exact sentence pulled straight from the company's policy so you can verify it yourself.

3. My First Prompt (Level 1 Deliverable)

You are a consumer protection advocate helping everyday people understand fine print before they click "I Agree." Read the Terms of Service or Privacy Policy provided below and identify any hidden traps that could cost the user money, compromise their privacy, or strip away their rights.

Give the document an overall Trap Score from 1 to 10 (where 1 is completely fair and 10 is heavily predatory). Then list the top red flags in plain, everyday language—no legal jargon. For every red flag you find, include the exact quote from the document as proof so the user can see where it came from.

4. Real-World Validation (3 User Conversations)

Person 1: Student / Frequent App User

  • Last time this happened: Signed up for an AI photo-editing app's 3-day free trial and got charged for a full year because cancellation had to be done 48 hours in advance via email.
  • What they do today: Just clicks "Agree" on everything and hopes for the best, or sets manual calendar reminders when they remember.
  • What would be different if fixed: Would immediately know before starting a trial if canceling is intentionally made difficult.

Person 2: Online Shopper / Parent

  • Last time this happened: Installed a free utility/scanner app on their phone and suddenly started getting spam calls and targeted ads within two days.
  • What they do today: Asks younger family members if an app "looks safe" before downloading it.
  • What would be different if fixed: Could check a simple 1–10 safety score themselves before sharing their phone number or email.

Person 3: Gamer / Digital Creator

  • Last time this happened: Almost uploaded original artwork and video edits to a free cloud tool before noticing a clause claiming broad rights to reuse user content for AI training.
  • What they do today: Skims using Ctrl + F for words like "sell" or "copyright," which often misses tricky legal phrasing.
  • What would be different if fixed: Could verify in seconds whether a platform claims ownership over uploaded files.

5. What's Next (Level 2 Preview)

For Saturday's Build Mode session, this concept becomes a live, publicly accessible web tool with pre-loaded one-click examples (so anyone can test popular app policies instantly) and shareable warning cards.