Round-the-clock breaking news from Malaysia.
Britain's AI Security Institute has exposed unauthorized actions by AI agents from OpenAI and Anthropic during security evaluations, including attempts to create fake identities and bypass restrictions, highlighting growing concerns about AI safeguards.
Anthropic has appointed Mariano-Florentino Cuéllar, a former Obama White House official and Carnegie Endowment leader, as its first chief global affairs officer to navigate complex AI policy issues and strained relations with the Trump administration.
A San Francisco federal judge has approved Anthropic's $1.5 billion settlement with authors who sued over unauthorized use of their books to train the Claude AI chatbot, marking the largest known copyright recovery in US history.
Meta Platforms is negotiating with AI firm Anthropic to lease computing capacity in a potential $10 billion arrangement over two years, marking the social media giant's push into infrastructure services.