An adversarial audit found that optimized fraud rings — fake identities rotating AI-generated content — could out-earn honest users six-to-one and quietly capture a third of all new money. Three ratified locks closed it. Below: you run the fraud ring. Toggle the locks and watch your business model die.
Near-identical content shares one shrinking payout — copies divide the pot instead of multiplying it.
Earning at scale requires posting a stake you forfeit if caught — and detection odds grow with ring size.
One lifetime identity per person. Renting yours to a ring risks losing it — so identities cost the ring dearly.
What to notice. Each lock alone only dents the heist — flip any two off and fraud pays again somewhere on the slider. It's the combination that kills the business model at every ring size: the similarity split starves revenue as the ring grows, the stake turns detection into a growing expected loss, and identity scarcity turns the ring's one "free" input into its biggest cost. Toy numbers, real principle. In the registered re-test against all three locks together, the heist didn't pay in a single tested case — with the honest caveat that this defense inherits the identity question (open item H-2), which only a live pilot can settle.