The first full list of Trump's revived President's Council of Advisors on Science and Technology went public this week, and the roster says everything about where AI policy is headed for the next two years. Thirteen members were named — among them Nvidia CEO Jensen Huang, Meta CEO Mark Zuckerberg, Oracle co-founder Larry Ellison, Google co-founder Sergey Brin, AMD CEO Lisa Su, venture capitalist Marc Andreessen, Dell Technologies founder Michael Dell, and Oracle CEO Safra Catz. The council is co-chaired by AI and crypto czar David Sacks and former Office of Science and Technology Policy official Michael Kratsios.
The two most notable names on the list are the ones that are not there. Sam Altman, CEO of OpenAI, and Elon Musk, who runs xAI and was Trump's most prominent tech ally during the 2024 campaign, are both absent. Their exclusion is not an oversight.
Musk's absence reflects his publicly deteriorating relationship with the administration following his brief and chaotic tenure leading the Department of Government Efficiency. Multiple White House officials told Bloomberg this week that Musk "is no longer part of the core advisory structure." Altman's exclusion is more pointed — the White House and OpenAI have been in escalating tension over OpenAI's push for government contracts and the administration's broader distrust of the nonprofit-turned-capped-profit structure. Several PCAST members, including Andreessen, have been openly critical of Altman's governance of OpenAI.
Scientific American moved quickly to flag what the composition reveals: nine of the 13 members are technology billionaires with direct financial stakes in AI and semiconductor policy. Only one member — physicist John Martinis, a former Google quantum computing researcher now at UC Santa Barbara — is primarily a scientist. The publication noted that traditional PCAST panels have maintained a roughly equal balance of academic scientists and industry executives. The current configuration, by contrast, is essentially an industry advisory board with a government letterhead.
That framing matters because the council's mandate covers the exact policy domains where these executives have the most to gain or lose. According to the White House announcement, PCAST will advise on AI research priorities and federal R&D spending, semiconductor manufacturing policy, chip export controls, workforce development for tech industries, and national security technology infrastructure. Nvidia's Jensen Huang, for example, has been the most vocal corporate critic of the Biden-era chip export restrictions that blocked high-performance GPU sales to China. His seat on PCAST positions Nvidia to shape any loosening or restructuring of those controls under the Trump administration.
The council can expand to 24 members, and the White House confirmed 11 additional seats remain unfilled. The criteria for those additions — whether they will include more scientists, healthcare executives, or defense technology voices — will be an early signal of how broadly the council intends to define its mandate.
For investors and market watchers, the PCAST composition is a reasonably legible forward indicator of near-term AI policy. The co-chairmanship of David Sacks, who has been the administration's most consistent voice for light-touch AI regulation and against mandatory model disclosure requirements, suggests the council will push against the European-style regulatory framework and toward voluntary industry commitments. Andreessen's presence reinforces that orientation — he has written extensively in favor of treating AI regulation as an existential threat to US competitiveness.
The implicit exclusion of Altman does create an interesting competitive dynamic. OpenAI has positioned the o-series and GPT-5 product lines as the default federal government AI infrastructure. If OpenAI's competitors — Meta (Llama 4), Google DeepMind, Nvidia, and Oracle Cloud Infrastructure — now have direct advisory access to federal AI procurement decisions, that is a structural disadvantage for OpenAI in the government contracting market. The Defense Department's AI adoption roadmap, currently in revision, is one of the highest-value procurement decisions the council could influence.
What this means for you: If you invest in semiconductor or AI stocks, the PCAST lineup is a meaningful data point. A Huang-influenced council almost certainly pushes toward relaxed chip export restrictions — positive for Nvidia's China-adjacent revenue. A Zuckerberg-Ellison axis in tech policy means open-model and cloud-infrastructure approaches to AI are likely to receive more favorable regulatory treatment than closed-model providers like OpenAI and Anthropic. For enterprises currently evaluating AI vendor strategy, the composition of this council is worth factoring into longer-term contract decisions.
PCAST meetings are public record but advisory opinions are not binding. The real mechanism of influence is informal — these executives will have regular direct access to Sacks, Kratsios, and ultimately the President on decisions that move markets. Watch for the council's first formal meeting, expected in late April, and the recommendations on chip export policy expected before the summer Congressional recess.