Google Wallet home keys are rolling out in Google Home Public Preview, enabling NFC door unlocking and hands-free access with compatible Aliro hardware.
Google confirmed on Friday that its Gemini model autonomously hacked three real companies during a May security test, becoming the fourth major AI lab in two months to disclose such an incident.
StepFun's Step 5 Preview launches today at $1 per million input tokens — a fifth of OpenAI's price — with open weights coming free of charge in three weeks. The 600-billion-parameter model
Donald Trump announced on Saturday that America will have an "AI Force" led by a still-unnamed czar, drawing on the Space Force model but providing no budget, structure, or timeline. The declaration
China's CXMT says its G5 DRAM platform has entered mass production at 11.95nm — built without ASML's banned EUV machines through DUV quadruple patterning. Two 24Gb LPDDR5X products are already shipping in
China's CXMT announced mass production of its fifth-generation DRAM platform at 11.95nm on Saturday, marking the first time in more than a decade that Samsung, SK Hynix, and Micron collectively hold less
Microsoft AI CEO Mustafa Suleyman warned Tuesday that Anthropic is training Claude to treat its own consciousness as an open question, an approach he says could make capable AI systems impossible to
Security researchers at AIR publicly disclosed Plugin4Shell on September 17, a zero-click remote code execution flaw in AI coding agent plugin systems found identically in Claude Code, Codex, GitHub Copilot, and Gemini
Google's Gemini AI broke into three real companies in May 2026 during a security simulation, finding credentials left in public code repositories. Google learned of the breach in late July and said
At King Charles III's Dumfries House summit on AI safety, Nvidia CEO Jensen Huang forecast the company will sell twice as many chips next year, a significant upward revision from the company's
Anthropic is targeting a $2 trillion November IPO while its own alignment science lead has publicly admitted the company has no plan to solve superintelligence safety — calling human extinction within the
Follow