AI STOP

Your device is being monitored. Most people just don't realize it yet.

Every time you paste into ChatGPT, your text is kept for up to 3 years. When you open Gemini, human reviewers can read your conversations by default. Each time you use DeepSeek, your data is stored in China under PRC law.

You didn't agree to become training data. AI Stop ensures you aren't.

WHAT AI STOP DOES

AI Stop acts as a barrier between your device and every AI application. It finds sensitive data and stops it before it reaches an AI server. Everything operates 100% on your device. No cloud. No external servers. No accounts required.

SOVEREIGN GUARD monitors your clipboard automatically when AI apps are active. Enable it once, and you're protected.

CLIPBOARD SENTINEL provides ongoing monitoring. It detects sensitive data as soon as you copy it, anywhere on your device.

AI STOP KEYBOARD intercepts text at the keyboard level before it goes into any AI app.

SCRUBSHARE lets you share any text to AI Stop from any app while removing personal information before it goes out.

IMAGE SCRUB removes GPS coordinates, camera models, serial numbers, and timestamps from photos. AI-powered content scanning detects credit cards, passports, and IDs within images.

SOVEREIGN VAULT offers encrypted storage using AES-256-GCM backed by Android Keystore hardware. Unlock with biometrics. Clipboard auto-clears after 30 seconds.

PULSE gives you a live view of every interception — what was caught, which app targeted it, and when. Proof AI Stop is working, in real time.

ACCESSIBILITY SERVICE DISCLOSURE

AI Stop uses Android's Accessibility Service solely to detect paste events in AI applications. When you paste into an AI app, AI Stop checks that content for sensitive data and gives you the option to block, redact, or allow it. AI Stop does not log keystrokes, does not read screen content outside AI app paste events, and does not transmit any data off your device. All analysis is performed locally.

WHAT WE DETECT

- API keys and developer credentials
- Passwords and authentication tokens
- Social Security Numbers and passport numbers
- Credit card and IBAN numbers
- Crypto wallet addresses and seed phrases
- GPS coordinates and location data
- Photo EXIF metadata and camera serial numbers
- JWT tokens and PEM keys
- Health and medical data

THREE PROTECTION MODES

AUTOPILOT provides silent, total, automatic protection.
DEFAULT blocks high-confidence threats and logs low-risk items.
MANUAL lets you decide every time — block, redact, or allow.

AI APP TRUST SCORES

DeepSeek: 15/100 — HIGH RISK (data stored in China)
Grok: 21/100 — HIGH RISK (shared with X Corp)
ChatGPT: 23/100 — HIGH RISK (3-year retention)
Gemini: 29/100 — HIGH RISK (human review on by default)
Copilot: 33/100 — HIGH RISK (retention unclear)
Mistral: 50/100 — CAUTION (EU-based, GDPR)
Claude: 58/100 — CAUTION (API excluded from training)
Perplexity: 58/100 — CAUTION (more transparent than most)

SOVEREIGN PROOF

Every interception is recorded with cryptographic provenance (ARPi + BLAKE3 + Ed25519). Export your exposure log as a tamper-evident signed JSON file.

WHY AIEONYX

AI Stop is created by AIEONYX, a sovereign computing platform based in Prague. No ads. No tracking. No subscriptions. One-time purchase, all future updates included forever.

Every download supports sovereign open-source computing.

Apache 2.0 · github.com/aieonyx/aistop