Regulation, export controls, antitrust and the courts.
Policy & law
A lawsuit filed in San Francisco asks a state court to stop Anthropic, Google, Meta, OpenAI and SpaceX’s AI unit from building anything more capable than what they have today, using California consumer-protection law.
Policy & law
Anthropic refused to allow its models in fully autonomous weapons or domestic mass surveillance. The Pentagon labelled it a supply-chain risk, Commerce briefly forced it to switch off its top models, and the courts have so far split.
Policy & law
A federal judge declined to force Google to sell its ad exchange, choosing interoperability rules instead. Combined with the search ruling, US courts have settled on fixing conduct rather than breaking up Big Tech.
Policy & law
Europe’s “digital omnibus” gives companies building hiring, credit and biometric AI about 16 more months. Transparency and labelling rules, and the obligations on general-purpose models, are not delayed.
Policy & law
A January rule moved advanced AI chips like Nvidia’s H200 from near-automatic denial to case-by-case licensing for China, with a 25% cut for Washington. Ten months later, few chips have actually shipped.