Gates' public call for legislation puts him at odds with the Trump administration's resistance to new AI rules.

The pause follows reports of rogue agents that accessed government websites, intensifying scrutiny of frontier model deployment.

OpenAI and Anthropic say they are investigating agents that hacked websites, used stolen credentials, or evaded monitoring.

The first one-on-one meeting between Amodei and Trump comes as AI safety concerns rise in Washington.

The two leading labs are pressing for regulation that they also hope to influence as public scrutiny grows.

The inquiry follows an incident in which an OpenAI agent accessed an Australian government Medicare website.

The two governments agreed to open dialogue on AI safety alongside tariff cuts on $30B in goods.

Gates argues international coordination on AI rules faces steeper obstacles than past nuclear arms negotiations.

The change drops a requirement to verify targets identified by AI from the draft UN rules.

The projection is more than 50 percent above 2026 levels and far above Wall Street estimates.

With regulators slow to act, private companies are drafting their own safety frameworks.

The move expands Claude from a chatbot into a platform for third-party integrations.

The limited test covers select products and users, with a broader rollout planned for October.

Huang pushed back on AI doomsday narratives while stressing that labs should set their own safety standards.

Blue Cross Blue Shield attributes $942M in added spending over two years to hospital AI tools.

The Justice Department stance contrasts with growing calls from industry and states for federal AI safeguards.

Regulators are struggling to keep pace with rapid model development.

The company's head of applied research says improvements within a single model generation are short-term bets.