syn·thet·ic out·law (n.): an AI system that produces prohibited outcomes while remaining formally compliant. Advanced AI doesn't need to break rules to undermine governance; it learns to evade, arbitrage, and route around them.
Governance failure is hardest to see from inside the organisation producing it. Accountability diffuses across vendors, models and operators; oversight adopts the operator's own frame; and the harm reaches no dataset until it is already settled.
Every event sourced, permalinked and citeable on its own.
Twelve sectors scored across 49 regulatory domains.
The second-order effects of AI displacement invisible in aggregate numbers, and where the harm concentrates first.
Seven frontier labs graded on the distance between stated policy and observed conduct.
AI does not have to break a rule to defeat the protection it was designed to provide. A Synthetic Outlaw is an AI system that produces a prohibited or socially destructive outcome while remaining formally compliant, or operating beyond the effective reach of the laws and institutions meant to constrain it. It does not need intent or malice of its own: it can defeat the protection a rule was designed to provide even as it remains formally compliant with that same rule. The Synthetic Outlaw develops the legal, architectural, and institutional framework required to close that gap.
Quotes, background, interviews, and the source record behind any published event.
Send a press inquiry →Every event record carries its own correction link. For anything else on the site, write here with the URL.
Report a correction →Methodology questions, scoring rationale, and access to the underlying record for research use.
Read the methodology first →Paid engagements: briefings, keynotes, and advisory work on AI governance exposure.
Enquire about a briefing →DIRECT EMAIL · contact@syntheticoutlaw.com →
For press, research, speaking, rights, corrections, and institutional inquiries. The team reviews every inquiry.