Researchers uncovered vulnerabilities in cloud-based password managers spanning recovery, encryption, sharing, and backwards compatibility.

The Pentagon is threating to cut a $200 million contract with Anthropic over a rift in how the Claude AI model should be used.