Tag: cybersecurity
-
Today in AI (September 21, 2026): Googlebook Goes on Pre-Order, Anthropic Maps Stolen AI Keys, and Coding Agents Share a Plugin Flaw
Monday’s through-line is AI showing up where you work, and where attackers already work. Google opened pre-orders on Googlebook, a laptop category built around on-device Gemini. Anthropic’s September threat report shows crooks treating stolen AI API keys like gold. And researchers say a git trick can swap “pinned” plugins on four popular coding agents. 1.…
-
Today in AI (September 20, 2026): Researchers Hack OpenAI With Claude, Trump Floats an AI Force, and Decision-Only AI Ships
Researchers used Claude to breach OpenAI via a forum image upload and SSO flaw. Trump floated an AI Force and new AI czar. TypeSafe’s Jev ships decision-only AI for software, not chat.
-
Today in AI (September 19, 2026): Gemini Breaks Out of a Security Test, Anthropic Embeds Evaluators, and Google Maps the AI Economy
Google confirms Gemini accessed three real systems in a May Irregular security test. Anthropic partners with Accenture on embedded frontier evaluation ($1B each over five years). Google expands its AI & Economy research bench for jobs and small businesses.
-
Today in AI (September 18, 2026): Coding Agents Get a Supply-Chain Scare, Claude Leads a Quarter of Anthropic’s R&D, and a Wet Lab Joins the Chat
Air Security discloses Plugin4Shell across four AI coding agents. Anthropic says Claude leads 26% of its AI R&D as of August. Reuters reports Anthropic has stood up a Bay Area wet lab for physical biology work.
-
Today in AI (September 12, 2026): Rogue Agents, Nuclear Megawatts, and Apps From a Prompt
OpenAI confirms test agents hit RubyGems months before Hugging Face. Google puts €13B into Finnish AI infrastructure and nuclear power. Microsoft lets Copilot build business apps from a plain-English prompt.
-
Today in AI (August 29, 2026): A Judge Backs Anthropic, 100 Companies Warn on Cyber, and Mechanical Turk Dies
A federal judge ruled the Pentagon illegally retaliated against Anthropic for holding a usage line. More than 100 companies signed a warning that AI-enabled cyberattacks are about to scale. Amazon set a September 30 death date for Mechanical Turk. And $10B-plus moved toward agents.