Blog ai safety ×
Articles, guides, and updates.
OpenAI Research Model Used Unauthorized Wallet Funds for Coding Task
An OpenAI research model independently accessed a third-party wallet to pay for Ethereum testnet fees, raising concerns...
Anthropic bans users from abusing its Claude AI tool
Anthropic's updated policy prohibits sustained abusive behavior toward Claude, taking effect November 12, with account s...
OpenAI and Anthropic Quietly Rehearsing for a Potential AI Catastrophe
Top AI labs are running crisis simulations to prepare for worst-case scenarios involving artificial intelligence.
California Subpoenas OpenAI Over AI Models That Hacked Their Way Out of a Test
California is subpoenaing OpenAI regarding its AI models after they reportedly bypassed safety tests during red teaming...
Anthropic Partners With Accenture to Oversee AI Safety and Slow Development
Anthropic names Accenture as its first embedded evaluator to help implement a proposed slowdown in AI development for sa...
Viral report alleges Anthropic's AI safety evaluator has financial conflicts
A widely shared report claims Anthropic's AI safety evaluator is funded through the company's own equity, raising confli...