Lab system cards and frameworks
60 of 329 documents, newest first
GPT-6 Cyber System CardSystem cardNov 2026140 paragraphsAt sourceOpenAIChatGPT Images 2.6 System CardSystem cardNov 202658 paragraphsAt sourceOpenAIGemini Omni 1.2 Flash Model CardModel cardNov 202648 paragraphsAt sourceGoogle DeepMindGPT-Rosalind-6 System Card AddendumSystem cardOct 2026107 paragraphsAt sourceOpenAIOpenAI’s Raising Concerns PolicyFrameworkOct 20268 paragraphsAt sourceOpenAIAddendum to GPT-6 Luna System Card: October 29, 2026 UpdateSystem cardOct 2026100 paragraphsAt sourceOpenAIAn updated Model Spec for more capable and beneficial AIArticleOct 202621 paragraphsAt sourceOpenAIAddendum to GPT-6 Astra System Card: GPT-6.1 TerraSystem cardOct 2026255 paragraphsAt sourceOpenAIBroad access through consistent standardsArticleOct 202614 paragraphsAt sourceOpenAIFrontier Safety Roadmap: archive of previous goalsFrameworkOct 2026100 paragraphsAt sourceAnthropicFrontier Safety RoadmapFrameworkOct 202661 paragraphsAt sourceAnthropicGPT-Live-1.1 system cardSystem cardOct 202651 paragraphsAt sourceOpenAIDistillation defensesArticleOct 202631 paragraphsAt sourceAnthropicDisrupting AI-enabled “false front” operationsArticleOct 202667 paragraphsAt sourceOpenAISystem Card: Claude Haiku 5.5System cardOct 2026144 pagesAt sourceAnthropicGPT-6 Sol and GPT-6 Luna: October 2026 updateSystem cardOct 2026152 paragraphsAt sourceOpenAITransparency HubArticleOct 2026793 paragraphsAt sourceAnthropicGemini 4 Argon Model CardModel cardOct 202614 pagesAt sourceGoogle DeepMindNano Banana 2.1 Model CardModel cardOct 20268 pagesAt sourceGoogle DeepMindDisrupting a coordinated model-distillation campaignArticleSep 202617 paragraphsAt sourceOpenAIAddendum to GPT-6 Astra System Card: GPT-6.1 SolSystem cardSep 2026262 paragraphsAt sourceOpenAISystem Card: Claude Sonnet 5.5System cardSep 2026148 pagesAt sourceAnthropicSystem Card: Claude Opus 5.5System cardSep 2026230 pagesAt sourceAnthropicBuilding standards for the next phase of AIArticleSep 202634 paragraphsAt sourceOpenAIIntroducing the Australian Youth Safety BlueprintArticleSep 20265 paragraphsAt sourceOpenAIAmazon's Frontier Model Safety Framework (September 2026)FrameworkSep 202611 pagesAt sourceAmazonOur framework for reporting model misalignmentFrameworkSep 202633 paragraphsAt sourceOpenAIThe AI policy window is open. We need to act.ArticleSep 202644 paragraphsAt sourceOpenAIChatGPT Images 2.5 System CardSystem cardSep 202653 paragraphsAt sourceOpenAIGPT-6 Astra System CardSystem cardSep 20261,010 paragraphsAt sourceOpenAISafety overview: GPT-6 AstraSystem cardSep 202612 paragraphsAt sourceOpenAISystem Card: Claude Fable 5.1 & Claude Mythos 5.1System cardSep 2026212 pagesAt sourceAnthropicEnterprise frontier safeguardsArticleSep 202628 paragraphsAt sourceAnthropicDetecting and countering misuse of AI: September 2026 (threat intelligence report)ArticleSep 2026154 pagesAt sourceAnthropicGemini 3.8 Audio (Live, Live Extended Thinking, Flash TTS, Flash-Lite TTS) Model CardModel cardSep 20267 pagesAt sourceGoogle DeepMindGemini 3.8 Flash Model CardModel cardSep 20268 pagesAt sourceGoogle DeepMindOpenAI – Hugging Face Incident Technical ReportArticleAug 202638 pagesAt sourceOpenAIDisrupting a new covert influence campaign from RussiaArticleAug 202622 paragraphsAt sourceOpenAIStrengthening democratic oversight in national securityArticleAug 202620 paragraphsAt sourceOpenAINew policy ideas for the Intelligence AgeArticleAug 202625 paragraphsAt sourceOpenAIClaude text watermarkArticleAug 202654 paragraphsAt sourceAnthropicRedacted Risk Report, August 2026ArticleAug 2026186 pagesAt sourceAnthropicAnthropic’s Responsible Scaling PolicyFrameworkAug 202698 paragraphsAt sourceAnthropicImproving Fable 5's biology safeguardsArticleAug 202619 paragraphsAt sourceAnthropicGPT-5.6 — August UpdatesSystem cardAug 2026187 paragraphsAt sourceOpenAIGemini 3.5 Audio (Live Translate, Transcribe, Transcribe Live) Model CardModel cardAug 20267 pagesAt sourceGoogle DeepMindGemini 3.7 Flash Model CardModel cardAug 20269 pagesAt sourceGoogle DeepMindDisrupting a Cambodia-based criminal scam operationArticleJul 202625 paragraphsAt sourceOpenAIInvestigating incidents in cybersecurity evaluationsEvaluation reportJul 202653 paragraphsAt sourceAnthropicGemini Robotics 2 Safety ReportEvaluation reportJul 202618 pagesAt sourceGoogle DeepMindPosition on open-weights modelsArticleJul 202612 paragraphsAt sourceAnthropicSystem Card: Claude Opus 5System cardJul 2026194 pagesAt sourceAnthropicTransparency Hub: System Trust and ReportingArticleJul 202653 paragraphsAt sourceAnthropicHard questionsArticleJul 202615 paragraphsAt sourceAnthropicGPT-5.6 System CardSystem cardJul 2026492 paragraphsAt sourceOpenAIOur approach to government and national security partnershipsArticleJul 20269 paragraphsAt sourceOpenAIGPT-Live System CardSystem cardJul 202645 paragraphsAt sourceOpenAIGemini 3.5 Flash-Lite Model CardModel cardJul 20267 pagesAt sourceGoogle DeepMindGemini 3.6 Flash Model CardModel cardJul 20267 pagesAt sourceGoogle DeepMindGemini Robotics ER 2 Model CardModel cardJul 20265 pagesAt sourceGoogle DeepMind