UK MPs call for AI regulator empowered to halt high-risk model releases
British lawmakers want a statutory AI regulator able to demand evidence and halt high-risk development. Their proposals are not yet law.
3 stories
British lawmakers want a statutory AI regulator able to demand evidence and halt high-risk development. Their proposals are not yet law.
NIST's ten initial agent identity tenets suggest the next AI bottleneck is not knowing which agent acted, but proving that the specific transaction was authorized.
A new systemic-risk research index finds major shifts when model ratings emphasize worst-case behavior. The lesson is to audit how the score was built.