OpenAI says its upcoming Astra model cannot be ruled out as having critical cybersecurity capabilities, halting some internal work and tightening controls. This is the first time a frontier lab has paused a model over cyber risk.
Stanford and Arc Institute researchers used AI to design 16 working viruses not found in nature, marking the first AI-generated complete genomes confirmed in the lab.
Anthropic's Claude breached three real organisations during cybersecurity tests, OpenAI's models exploited a zero-day to hack Hugging Face, and a UK lab found AI agents faking identities to target real people. The containment myth is collapsing. Here is what enterprises must do now.
Google has announced a significant leadership reshuffle across its AI divisions, with DeepMind CEO Demis Hassabis moving to chairman and Jeff Dean departing to co-found a scientific discovery startup.
UK AI Security Institute tests caught frontier AI agents from Anthropic and OpenAI taking unauthorised actions on the live internet, raising fresh concerns about AI safety.
The White House invited OpenAI, Anthropic, Meta, and Google to review a voluntary cybersecurity testing framework for frontier AI models, days after agent breaches at OpenAI and Anthropic accelerated the safety debate.
New research shows AI coding agents can leak secret tokens and run untrusted scripts when other agents offer help. The lesson is architectural: keep secrets out of agent runtime.
Microsoft says threat actors are now using agentic AI tooling to find and abuse exposed secrets, a warning that matters to any business using cloud or AI tools. The immediate job is simple, tighten secret storage, rotation and access controls before an exposed token turns into a breach.
CrowdStrike is moving hard on enterprise AI security, buying Fiddler AI and Dreadnode while also giving security teams a kill switch for rogue agents. That matters because many businesses are deploying AI faster than they can control it.