Abu Dhabi’s AI-native government runs on one app, TAMM, that renews your ID and pays your fines. The catch is a model no democracy can copy.

ASML says Intel is using its $400M High NA EUV tool to make Panther Lake laptop chips. It is a first for the technology, and earlier than planned.

Albanese announced an Office of AI, net-generator rules for data centres, and tough copyright language. None of it is law yet.

Gold Eagle is the White House’s new AI-backed clearinghouse to pool, rank, and coordinate fixes for cyber flaws across US critical infrastructure.

Anthropic safety hiring now targets nuclear, chemical, and biological harm, staffing analysts to stop its AI ever helping anyone build a weapon.

Lucid shares plunged after a report said AlixPartners was reviewing Chapter 11 and going-private options. Lucid denied the claims. It closed down 16% at $4.62.

OpenAI’s first io-designed device is a screen-free smart speaker with a camera, a battery, and parts that move on their own, Bloomberg reports.

The Delaware AIC would give an AI agent its own legal entity to sign contracts, own property, and be sued, all inside a supervised regulatory sandbox.

Judge Noël Wise dismissed the iCloud CSAM class action with prejudice, ruling no federal law compels Apple to scan and that Section 230 covers the claims.

A researcher proved Grok Build CLI sent full repos, 27,800x more data than the task needed, to a Google Cloud bucket. xAI had said no code was transmitted.

Upwind traces multiple compromised AsyncAPI npm packages to a coordinated supply chain attack targeting software release pipelines and publishing identities.

Anthropic is funding Amii, Mila, Vector Institute, and five other Canadian institutions. Canada ranks second worldwide in per-capita Claude usage behind the US.