Leadership1 publisher3 min readPublished
Trump and Vance wave off the federal AI rulebook Altman says he would welcome
OpenAI says it would welcome slower frontier development and consistent federal safety requirements for it. Trump answers that a strong president is the guardrail AI needs, and Vance calls the industry's request a Trojan horse.
The Board Room · Leadership desk
What happened
- Sam Altman said OpenAI would welcome a slower pace of AI development so that the technology's capabilities do not get ahead of alignment and monitoring.
- Altman said late Sunday that the company would welcome a federal framework setting consistent safety requirements for frontier AI.
- Trump posted on Monday that the only control or guardrails AI needs is a strong and smart president, and that the United States has that in spades.
- Vance said Monday that the administration is concerned about the technology but that any regulation which comes needs to be done smartly.
- Jacob Coxon, a former safety researcher at Anthropic and OpenAI, said in a social media post that he was resigning from Anthropic and that neither company is acting responsibly.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- exposure The second failure mode Altman named, too much concentration of power, describes the position his enterprise customers already occupy, and the company deciding how much of it is acceptable is the company they depend on.
- contradiction OpenAI asks for consistent federal rules and the vice president treats labs asking to be regulated as a Trojan horse, so the same request is heard as risk management on one side of the table and as moat-building on the other.
- constraint An internal AI policy drafted this quarter cannot use a federal frontier safety requirement as its floor, so the floor is whatever the board sets and whatever the vendor has published.
- precedent If voluntary pacing commitments are the only ones in force while Congress is away, later statutory language will most likely be drafted around what the leading vendor has already agreed to do.
One company cannot grant a slower industry pace, and Altman's own sentence names what granting it would take. "But we do not believe we need to wait for an anti-trust exemption or legislation to begin the work of providing this confidence," he said [4]. Competing labs agreeing to hold capability releases back is coordination among rivals, and an exemption is the instrument that would make that coordination safe to do in the open. Altman said OpenAI will proceed without one.
Vance is skeptical of the request itself. He said he feels "a little bit weird about the fact that you have so many Frontier AI tech companies kind of coming to the government and begging the government to regulate them", and that "it feels a little bit to me like a bit of a Trojan horse" [11]. The president was blunter about who gains from restraint, saying there is a "SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China" [9]. Trump also framed the technology as a race, telling reporters on Sunday: "Whoever wins AI -- and we're leading by a lot. Whoever wins AI wins." [7]
House Democratic Leader Hakeem Jeffries wants the pace question settled by statute. "We should take decisive action now so that we can slow down, as the CEOs have recently acknowledged, slow down the pace of development in order to protect the American people and ensure that AI is proceeding safely," Jeffries said on the ABC News program This Week [16]. Lawmakers from both chambers return to Washington this week for the final legislative push before the November midterm elections [17]. A frontier safety statute passes in that window or waits until after the elections [21]. Members of both parties have already demanded information from OpenAI and opened investigations [18]. Investigations produce documents on a committee's schedule. A procurement officer cannot cite one in a contract.
The sharpest claim in the account came from someone who left. "The people building AI earnestly believe that it could kill us all by the end of the decade," Jacob Coxon wrote [14]. ABC News carries that as one departing researcher's characterisation of two companies, sourced to him alone. Altman's framing is narrower. He named an "unacceptable" loss of control and a world with "too much concentration of power" as the two ways progress goes very badly [19], and said "If an extraordinarily powerful AI is used by one person or company to impress their worldview onto everyone else, the results could be extremely dystopian" [20].
For anyone signing a deployment this quarter, the live question is which document governs, and the answer does not depend on whether Vance is right about motives. Waiting for a federal baseline costs deployment time against a date nobody has published. Proceeding means the limits on model autonomy and logging are the ones the contract sets, alongside whatever the vendor publishes and can revise. The outside check Altman named is review by others: "Consistent rules to manage frontier risk so that we can maximize the benefits are a good idea (and we are excited by ideas like independent auditors)," he said [5]. The account leaves the scope of such audits, and the date they would start, open. Altman said avoiding both threats requires "walking a narrow middle path" [2].
What to watch
- Whether any other frontier lab commits in writing to the pacing Altman described, or OpenAI stands alone on it.
- Whether the White House turns Vance's 'regulate smartly' into a specific proposal or leaves the drafting to the labs.
- Whether OpenAI's release schedule slows in a way an outside party can measure.