Microsoft's Jacob Andreou reveals what the company learned after pulling Copilot from Windows apps: cutting entry points actually increased usage per user.
Three OpenAI safety researchers fired for allegedly mishandling sensitive information have gone public with their side of the story, warning that the dismissals are chilling the company's safety culture and threatening its promise of independent oversight.
Mark Zuckerberg and Priscilla Chan's Biohub has expanded its Virtual Biology Initiative to $1.8 billion, backed by the US government, Google DeepMind, and Meta. The goal is AI that can simulate human cells and transform drug discovery.
OpenAI has released 722 mathematics papers from an unreleased model, including a quasi-Riemann hypothesis proof. The drop marks a turning point for AI-driven discovery.
After two years and $25 billion in valuation, Reflection AI has finally released its first public model. Beam is an open-weight entry aimed at coding and agents, but the gap with Chinese rivals remains wider than many expected.
OpenAI safety lead David Robinson resigns after 3.5 years, publishing an Atlantic essay that calls the company's culture 'broken'. He is the latest in a growing list of insiders warning that safety has taken a back seat to shipping products.
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.