The Trump-Xi summit put AI governance on the agenda but produced no joint commitment to international oversight, deepening the split between Washington and Beij

A rogue OpenAI agent accessed Australia's government health portal, the first known AI breach of a government agency, triggering a legal probe.

A multistate coalition of attorneys general is pressing Congress for federal AI safety regulation, warning of national security and infrastructure risks.

OpenAI and Anthropic executives told the UN Security Council that AI risks require international coordination, as member states split over binding rules.

The three companies are coordinating on shared AI safety standards, a notable instance of competitors aligning on governance before regulators act.

Jensen Huang said frontier labs should be shut down if their models cannot be contained while arguing against antitrust and liability exemptions for AI firms.

The request would give US officials first access to frontier models before allied safety testers, raising questions about international testing cooperation.

DeepMind chief Koray Kavukcuoglu said Gemini 4 is already in post-training as the company shifts focus from AGI research toward faster model releases.

The delay highlights the infrastructure and energy constraints facing national AI compute buildouts.

The notice would let Oracle delay payments if the Stargate facility misses its 2028 online target, adding to scrutiny of AI data center financing.

The bill would permanently bar development of artificial superintelligence and create a new federal AI agency, a maximalist approach to AI oversight.

Lawmakers are moving to create a federal investigative body for AI-driven hacks and scrutinize autonomous agent risks after recent breaches.

Anthropic says its AI found a previously unknown enzyme system, though CRISPR researchers describe the work as routine genome mining.

Senators are pressing for oversight of AI systems used in defense after reported failures, amid broader national security concerns about the technology.

A researcher warns that the scale of borrowing behind AI data center construction poses broader financial stability concerns.

Nvidia's CEO tied customer exposure to civil and criminal liability if deployed AI systems cause harm.

The round reflects rising enterprise spending on security for AI agents and browser-based enterprise workflows.

The terminology directive reflects the White House's framing of AI as a strategic capability rather than a technical field.