How ECE AI works

Three pillars turn a single piece of harmful synthetic media into an understandable, evidence-backed trail — and a responsible plan for what to do next.

Trace

ECE AI organizes publicly observable occurrences of a piece of media into a chronological trail — starting from the earliest publicly observable occurrence, not an assumed 'original uploader.'

In this prototype, that trail comes from a fixed demo dataset rather than a live search.

Verify

Every finding is sorted into one of three buckets: evidence that's been observed, reasonable inference drawn from that evidence, and genuine uncertainty that remains.

This separation is the core of ECE AI's media & information literacy approach — no single AI verdict is presented as final truth.

Respond

The evidence trail is turned into a structured, exportable response pack — timestamps, public source references, confidence explanation, and limitations — ready to hand to an authorized responder.

The pack always includes responsible next steps: preserve evidence, avoid resharing, use official reporting tools, and seek appropriate support.

Product positioning

ECE AI is not

A generic deepfake detector
A people-search engine
A doxxing tool
A takedown bot
A law-enforcement replacement

ECE AI is

An evidence-intelligence workflow
A propagation-tracing prototype
An MIL-oriented verification assistant
A responsible response preparation tool

See it in action

Walk through the full demo flow with a fictional case, end to end.

Start investigation