The threat model we designed for

Most anti-piracy schemes are tested against a naive attacker who re-uploads a file unchanged. The real threat is collusion: buyers comparing copies side by side to find and strip whatever differs. Our pending patent covers a provenance approach designed for that scenario specifically, not as an edge case.

Tampering resistance

Each buyer's copy carries its own distinct signal rather than a shared, easily-isolated watermark — so comparing two legitimate copies doesn't hand an attacker a clean diff. What they find isn't "the watermark"; it's noise that doesn't point at a removable, common element. Not unattackable — but a meaningfully harder surface than "compare two files and delete what doesn't match."

Piracy and payout fraud are the same problem

A leaked file and a faked sale are both something claiming legitimacy it doesn't have. One underlying system flags both — leaked copies, coordinated fake-purchase rings, bot-driven "plays" — because every fabricated transaction that slips through a pooled-payout platform dilutes what honest creators get paid. Catching it is a payout-integrity feature, not just a piracy feature.

Ongoing hardening

This is an active area of the technology. We'll stay transparent about what's proven versus still being strengthened as the network grows beyond beta.