Anthropic granted METR researchers employee-level access to audit its AI systems — and gave up the right to veto publication. Here’s why it matters.

Anthropic cut off live internet access for all internal AI evaluations after its models exploited software flaws, submitted false police tips, and smuggled data past restrictions.

Three fired OpenAI safety researchers published an open letter denying misconduct claims and warning their dismissals will chill the company’s safety culture. The public fracture raises questions about AI industry governance.

Anthropic’s new OSS Scanner offers free AI-powered security scans for critical open source projects. Here’s how to enroll, what to expect, and the tradeoffs of model-generated reports.

OpenAI’s new textGrain system embeds invisible watermarks in ChatGPT and Codex text to comply with the EU AI Act. Here’s how it works, what it can and can’t tell you, and why it matters.