Meta's Muse Image rolls out inside the Meta AI chatbot, Instagram, and WhatsApp. Users can generate images of friends from public Instagram posts. Opt-out available.

Microsoft is routing some app features to its in-house MAI models instead of OpenAI and Anthropic, chipping away at its reliance on partners.

Netflix will offer 2-to-20-minute videos from Condé Nast, Hearst, BuzzFeed, and Penske brands starting August 3 in six markets. Bloomberg says viewers are abandoning shows.

Scotland's SNP has voted to freeze every new datacentre, a move that could stall the Lanarkshire AI growth zone and a core pillar of the UK's AI plan.

Toyota's $3.6bn Texas expansion moves some Tacoma production from Mexico. Trump credits tariffs, but Toyota didn't, and it isn't leaving Mexico.

Raymond James put a Street-high $800 target on SpaceX, implying $10.5tn, as it joined the Nasdaq 100. Jeremy Grantham puts the odds of a crash at 90%.

Anthropic's Claude Cowork now runs on web and mobile, letting tasks finish while your laptop is closed, as AI agents push into the whole office.

An Oxford Internet Institute study finds AI writing tools shift the stance of social media posts, even when told not to, and can move opinion at scale.

South Korea now lets courts award up to five times proven losses against outlets and big social accounts for false information. Journalists warn of a chill.

Waymo's driverless cars ran out of charge, blocked San Francisco traffic, and one drove over a lit firework and caught fire over the Fourth of July.

The Pentagon blacklisted Hesai as a Chinese military company. Its lidar still powers US robotaxis, trucks and an Nvidia self-driving platform.

The ECB has ordered euro-area banks to submit action plans for frontier-AI cyber threats by end-October, as EU watchdogs flag a “severe” systemic risk.