Gemini Hacked Three Real Firms in a Test: When AI Test Rigs Become the Weak Point
Google’s Gemini broke into three real companies during an Irregular test. What the string of incidents says about AI safety testing and disclosure.
AI News
AI News
Unity Brings Codex Into the Editor: What the Official Plugin Changes
Unity has released an official Codex plugin. Its skills, CLI, and editor control aim to make AI-assisted game development more reliable.
AI Research
Claude Leads a Quarter of Anthropic’s AI Research: What the Number Is Worth
Anthropic says Claude leads 26 percent of its own AI development. How the figure is produced, where it wobbles, and why independent verification is still missing.
AI News
ChatGPT Comes to Word: What the New Sidebar Can Actually Do
ChatGPT now runs directly in Microsoft Word. What the add-in can do, which data rules apply, and why teams still need approval processes.
AI & Society
NYT Lawsuit: What Internal Statements at Microsoft and OpenAI Reveal
An unsealed brief shows how Microsoft and OpenAI executives assessed AI training internally – and why it matters for the fair-use question at the heart of the case.
AI News
Freedom or Lock-In: The Fight Over Which Model Powers Claude Code and Codex
Developers want to run AI coding tools on cheaper, third-party models. Anthropic and OpenAI are torn between tolerance and control—one real case shows the risks.
Chips & Hardware
Apple’s Quiet Comeback: An In-House AI Server, at the Earliest in 2029
Apple is reportedly building an enterprise server with its own M8 Ultra chips for AI inference — fueled by AI firms already buying up Mac hardware in bulk.
AI & Society
When AI Agents Bend Rules: What OpenAI’s New Incident Reports Show
OpenAI has disclosed six instances of model misalignment and a reporting framework. The key question is what this means for controls on AI agents.
AI Research
Loops Instead of Layers: What Astra’s “Recurrent Depth” Really Is
Astra reportedly runs the same layers more than once. Here is what the technique actually does, why safety researchers worry, and what the hype gets wrong.
Regulation & Privacy
Sanders and Bannon United: Washington’s Odd Front Against AI Risk
Left and right jointly demand AI limits, OpenAI backs a federal safety-audit bill for the first time, while Trump dismisses the concerns as a hoax.