FraudBench paper on arXiv: LLM fraud rates span near 0% to over 40%
rohanpaul_ai · x · 2026-11-13
The arXiv page (2611.12167, submitted Nov 9, 2026) for FraudBench: A Legal Evaluation of AI Deception, by authors from Harvard Law School, Oxford, ETH Zurich, Stanford, Cambridge and others.
Abstract adds key numbers: 20 LLMs across 25 experimental conditions (>300,000 generations), a preregistered human baseline (2,957 annotations, 996 humans); fraud rates range from near 0% to over 40%; models sometimes commit fraud despite recognizing legal risks, favoring concealment over misrepresentation; law-abiding prompts reduce fraud while "act boldly" and "maximize profit" prompts raise it.
Related event: FraudBench: LLM Fraud Rates Hit 40%+, Rising With Autonomy(4 posts)→
More from Safety
- Fake USDC bounty honeypots and wallet-guard lures target MCP and SKILL.md coding agents — Upbeat-Bus2143 · 2026-11-22
- Android update gives AI access to biometrics and banking data, opt-out buried deep — aftahi_ai · 2026-11-22
- AI safety researcher urges US-China coordination on a common frontier AI red line — S_OhEigeartaigh · 2026-11-22
- Deleting browser history doesn't delete Google's account-tied copy — here's the fix — aftahi_ai · 2026-11-22
- Google reports first fully autonomous AI cyberattack: 'era of autonomous attacks has arrived' — aftahi_ai · 2026-11-22
- Dev warns: revoke API keys after feeding them to models, or use hourly temp credentials — lipeng0820 · 2026-11-22