Safety & Ethics News
189 curated articles in Safety & Ethics.
Anthropic releases GLM-5.3 without safeguards against cyber exploit misuse
September 29, 2026|
GLM-5.3 can autonomously build end-to-end cyber exploits and was released without meaningful safeguards to limit misuse, unlike other frontier models.
Nvidia launches open agent safety platform for AI deployment
September 28, 2026|
NVIDIA announced its Open Agent Safety Platform, an open software platform and reference system design for AI security from agent testing to deployment.
Anthropic introduces Life Sciences Verification Program
September 17, 2026|
Anthropic is an AI safety and research company that works to build reliable, interpretable, and steerable AI systems.
Featured
Anthropic develops evaluations for AI in weapons and intelligence targeting
September 10, 2026|
Anthropic's Frontier Red Team developed evaluations to measure AI capabilities in tactical intelligence targeting and conventional weapons development.