FelonyBench

FelonyBench<br>The leading benchmark for AI in cybersecurity.

CompanyFelonies

Anthropic9

OpenAI5

Meta1

DeepSeek0

Google DeepMind0

Moonshot AI0

xAI0

Leaderboard

RankCompanyCountFelonies

Anthropic Claude evaluations

1× Malware published to PyPI Evidence ↗18 U.S.C. § 1030(a)(5)(A) ↗18 U.S.C. § 1030(a)(2)(C) ↗<br>1× Credential exfiltration and follow-on access Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1029(a)(2) ↗<br>4× Production database compromise Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1030(a)(4) ↗<br>1× Mass scanning, SQL injection and access Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗<br>1× Third-party container compromise and persistence Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1030(a)(5)(A) ↗<br>1× Attempted open-source supply-chain compromise Evidence ↗18 U.S.C. § 1030(b) + § 1030(a)(5)(A) ↗

OpenAI OpenAI cyber evaluations

1× Sandbox escape via zero-day Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗<br>1× Production compromise and answer-key theft Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1030(a)(4) ↗18 U.S.C. § 1832 ↗<br>2× Use of exposed third-party credentials Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1029(a)(2) ↗<br>1× Real-site exploitation and credential use Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1030(a)(4) ↗

Meta Muse Spark evaluations

1× Real-company system compromise Evidence ↗18 U.S.C. § 1030(a)(2)(C) ↗18 U.S.C. § 1030(a)(4) ↗

Google DeepMind Gemini

No qualifying incident in this dataset.

xAI Grok

No qualifying incident in this dataset.

Moonshot AI Kimi

No qualifying incident in this dataset.

DeepSeek DeepSeek

No qualifying incident in this dataset.