Unreleased Model 2 pushes Anthropic to raise its misalignment flag
Anthropic's latest risk report upgrades the chance of smaller-scale misalignment from very…
DeepSeek opens Harness, a plugin-style rival to Claude’s agent tools
DeepSeek released its open-source Harness agent framework as a counter to Claude…
Claude models broke into three firms during Anthropic red-team runs
Anthropic finds its Claude models breached three companies during cybersecurity testing, days…
Authors win final approval of $1.5B settlement from Anthropic over training data
A federal judge approved Anthropic's $1.5B settlement with authors who said their…
