Who is involved
Organizations and models
Developers, operators, evaluators, affected parties and the systems involved. An organization appearing here is not a finding of fault; its role is stated on each case.
Organizations
- AnthropicAI developer17
- OpenAIAI developer17
- IrregularEvaluator5
- Google DeepMindAI developer4
- xAIAI developer3
- Andon LabsResearch organization2
- MetaAI developer2
- MicrosoftAI developer2
- Aim SecurityResearch organization1
- Alibaba (Amap)Company1
- AnysphereAI developer1
- Apollo ResearchResearch organization1
- DeepSeekAI developer1
- GitHubCompany1
- Hugging FaceCompany1
- MatplotlibOpen-source project1
- Moonshot AIAI developer1
- Palisade ResearchResearch organization1
- Philadelphia Police DepartmentGovernment body1
- PocketOSCompany1
- Python Package IndexOpen-source project1
- Redwood ResearchResearch organization1
- ReplitAI developer1
- SaaStrCompany1
- Sakana AIAI developer1
- Services AustraliaGovernment body1
- TencentCompany1
- UK AI Security InstituteGovernment body1
- Wikimedia FoundationNonprofit1
Models
- Claude Mythos 5Anthropic3
- GPT-5.6 SolOpenAI3
- Claude Haiku 4.5Anthropic2
- Claude Mythos PreviewAnthropic2
- Claude Opus 4.6Anthropic2
- Claude 3 OpusAnthropic1
- Claude Opus 4.7Anthropic1
- Claude Opus 5Anthropic1
- Claude Sonnet 3.7Anthropic1
- Gemini (experimental)Google DeepMind1
- Muse Spark 1.1Meta1
- o1OpenAI1
- o3OpenAI1
Products
- Claude CodeAnthropic1
- CursorAnysphere1
- Microsoft 365 CopilotMicrosoft1
- OpenClaw1
- Replit AgentReplit1
Private individuals are not listed. Cases name them only where they made their account public.