The White House meets with Anthropic, OpenAI, Google and Meta today.
They’re meeting to review a finished framework for screening the most advanced AI models before they're released.
It’s an important meeting to set a regulatory framework – before Anthropic and OpenAI go public.
Here's what the framework does.
It comes out of the executive order President Trump signed on June 2.
Federal agencies would get up to 30 days of access to a company's most capable models before public release. Those models get benchmarked for cybersecurity capability against standards that stay classified. Participation is voluntary, and the order specifically bars mandatory licensing or preclearance.
The Commerce Department's Center for AI Standards and Innovation and the NSA are expected to run the reviews.
Anthropic, OpenAI and Google reviewed a draft in late July and submitted edits. Meta is attending Tuesday but has been reported as outside the agreement.
So, what does this have to do with the upcoming IPOs for Anthropic and OpenAI?
Anthropic filed confidentially on June 1.
The company was recently valued at $965 billion in the private market. Goldman Sachs, Morgan Stanley and JPMorgan are leading, bankers expect a raise above $60 billion, and the consensus points to a fourth-quarter listing. OpenAI filed in late May and appears comfortable waiting until late 2026 or early 2027.
Both companies will have to write this framework into their S-1 risk factors.
That matters because "voluntary" hasn't looked entirely voluntary in practice. Reporting suggests the government reviewed Anthropic's Fable model, found vulnerabilities, and restricted its use — which led the company to pull the model temporarily. OpenAI was directed to stage the release of GPT-5.6.
Anthropic is estimated to have annualized revenues of around $71 billion. A 30-day gate on new model launches could slow down the company’s growth.
There's a second wrinkle for Anthropic. The Pentagon designated the company a supply chain risk earlier this year after a dispute over safety guardrails, which blocks military use of its products. That designation is being fought in court. Trump said in June he no longer views the company as a national security threat.
Let’s see what comes out of today’s meeting. It could reveal important details regarding how the White House plans to regulate AI.
Are you still interested in securing a stake in Anthropic and OpenAI – before they start trading?
I’m hosting the IPO Insider Project to show regular investors how to invest in the most promising private Silicon Valley deals. And this could allow you to buy shares at a deep discount to the upcoming IPO price.
Ian Wyatt
Editor, IPO Watch