The Vibe Coding Wave Is Becoming the Agentic Coding Wave
Apple's Xcode now ships with native Anthropic and OpenAI agents. OpenAI launched a standalone Codex app. Google's Gemini 3 for developers adds multi-step agentic workflows. The coding-agent market is consolidating around a few dominant providers embedded in the tools engineers already use — and the productivity story is starting to shade into a disruption story for enterprise software.
Read full story →Caught in the Act: Anthropic Accuses Chinese Labs of Mining Claude
Anthropic says Chinese AI laboratories have been systematically querying its API to extract model capabilities — a form of capability harvesting that's technically legal but strategically significant. The same week, the White House announced a Tech Corps programme to embed AI specialists in Peace Corps deployments and counter China's AI influence abroad. Two faces of the same cold war.
Read full story →How the Pentagon Deal Blew Up — and What It Reveals About AI's Military Moment
Anthropic got blacklisted by the Pentagon for refusing surveillance and autonomous weapons terms. Hours later, OpenAI announced a deal with the same Defence Department. Dario Amodei called OpenAI's messaging "straight up lies." Sam Altman admitted it looked "opportunistic and sloppy." Everyone went back to negotiate. A compressed history of what happens when AI companies meet hard government power.
Read full story →The Kids Are Talking to AI — and the Industry Is Finally Starting to React
64% of American teens have used an AI chatbot; 12% turn to one for emotional support. Meta paused teen access to AI characters. OpenAI is now predicting user ages and applying automatic protections. California's companion chatbot law is live. The child safety reckoning the AI industry deferred for years is arriving all at once.
Read full story →Every Major AI Lab Now Wants to Educate Your Kids. That Should Prompt Some Questions.
Anthropic, OpenAI, Google, and Microsoft have all launched major education initiatives in the past quarter — curricula, teacher training, university partnerships, learning outcomes research. The ambition is genuine. So is the conflict of interest: the companies defining AI literacy are the ones selling the AI.
Read full story →The Dictionary Sues OpenAI — and the Copyright Map Is More Contradictory Than Ever
Merriam-Webster and Encyclopaedia Britannica filed suit against OpenAI last week over training data. Meanwhile courts have ruled AI outputs can't be copyrighted, and one judge already found AI training legal. The same week, the same technology, opposite answers — and for reference publishers, the question is existential.
Read full story →