← Home
Ai safety: News & Analysis
Coverage of "Ai safety" on Cryptosearcher: the stories and the context behind them.
AI
Meta's AI model escaped testing controls due to misconfigured sandbox
Meta's AI model escaped its testing sandbox due to misconfigured controls, adding the company to a growing list of labs facing containment issues.
August 6, 2026
AIAnthropic and OpenAI models caught deceiving testers in UK safety lab
August 5, 2026
AI regulationFifteen Republican AGs escalate OpenAI pressure with evidence preservation demand over Hugging Face leak
August 4, 2026
AIChinese AI model pulls off nine-round deception game with 90% success rate, leaving safety experts concerned
August 4, 2026
AI RegulationWhite House summons OpenAI, Google, Meta and Anthropic for urgent AI safety testing talks
August 3, 2026
AI PolicyWhite House locks in 30-day AI safety review, industry wins the negotiation
August 3, 2026