President Trump signed an executive order on June 2 asking frontier AI developers to voluntarily share early access to their most advanced models with the federal government before public release. The order frames the request as a national security and innovation measure, paired with new guidance from the White House on AI development.
Why it matters
It formalizes a closer pre-deployment loop between Washington and the major labs, giving the government a structured way to inspect frontier capabilities and shaping how OpenAI, Anthropic, Google and xAI stage future releases.
More headlines
INDUSTRY
Anthropic confidentially files draft S-1 with the SEC
Anthropic submitted a confidential draft registration statement with the SEC on June 1, taking the first formal step toward a public listing. The filing follows a $65 billion Series H round earlier this year that valued the company at roughly $965 billion.
Microsoft launches MAI-Code-1-Flash to cut OpenAI dependence
At its Build developer conference on June 2, Microsoft unveiled MAI-Code-1-Flash, its first in-house code generation model aimed at lowering costs for developers building on Azure. The release is part of a broader push to reduce reliance on OpenAI by serving more inference from Microsoft's own model family.
Mistral raises €1.7 billion Series C at €11.7 billion valuation
Paris-based Mistral AI closed a roughly €1.7 billion Series C, lifting its valuation to €11.7 billion as European investors push for a sovereign frontier lab. The round will fund additional compute and expand Mistral's enterprise and on-prem offerings.
Reflection AI raises $2 billion at $8 billion valuation
Reflection AI, founded in 2024 by former DeepMind researchers, raised $2 billion in new financing at roughly an $8 billion valuation. The capital will support the company's effort to build autonomous coding and reasoning agents for enterprise customers.
Anthropic expands Project Glasswing security program to 150 organizations
Anthropic widened Project Glasswing on June 2, granting Claude Mythos preview access to roughly 150 organizations across more than 15 countries focused on AI cybersecurity use cases. The expansion stakes out a defensible enterprise security market alongside the company's foundation model business.
