Built privacy-first
Every design decision in ECE AI starts from what the product must never claim or access — not just what it can do.
No unnecessary sensitive-media storage
Demo submissions in this prototype are processed in-session only and are not retained or stored.
Fictional demo content only
No real victims, no real intimate images. Every case, handle, and timestamp shown is simulated.
No real-world identity claims
ECE AI never claims to identify the real person behind an anonymous account — in the demo or in its intended design.
No private-account access
The product never claims access to private accounts, direct messages, or non-public platform data.
No unauthorized surveillance
No IP addresses, device fingerprints, telecom records, or law-enforcement databases are accessed.
No automatic takedown claims
ECE AI does not claim to remove content from the internet. It prepares evidence for authorized responders to act on.
Who evidence should be shared with
Evidence produced by ECE AI is intended for the case owner and appropriate authorized responders only — such as platform trust & safety teams, victim support services, or a relevant legal or cybercrime authority. It is not intended to be shared publicly, used to identify or contact suspected individuals, or used as grounds for independent investigation or retaliation.
Language we use carefully
- “Earliest publicly observable occurrence”
- Used instead of “original uploader,” because ECE AI can only speak to what became publicly visible first — never who truly created or first possessed the media.
- Evidence vs. inference vs. uncertainty
- Every claim is labeled as one of these three, so a reader can judge how much weight to give it.