Trump's PCAST 2026: Zuckerberg, Huang, Ellison Shape AI
On March 25, 2026, President Trump appointed 13 members to the President's Council of Advisors on Science and Technology (PCAST), including Meta CEO Mark Zuckerberg, Nvidia CEO Jensen Huang, Oracle's Larry Ellison, Google co-founder Sergey Brin, AMD CEO Lisa Su, and Marc Andreessen. The panel is co-chaired by AI czar David Sacks and OSTP director Michael Kratsios, and can expand to 24 members. First policy recommendations are expected within 90 days. Neither Elon Musk nor Sam Altman was appointed.
Washington has always understood that the most important meetings are the ones that never appear on the official calendar. On March 25, 2026, that reality was made unusually explicit.
President Trump appointed Meta CEO Mark Zuckerberg, Oracle Executive Chairman Larry Ellison, and Nvidia CEO Jensen Huang to the President's Council of Advisors on Science and Technology — a panel that will weigh in on AI policy and other issues, the White House announced Wednesday.
The initial 13 appointees include Marc Andreessen, Sergey Brin, Safra Catz, Michael Dell, Jacob DeWitte, Fred Ehrsam, Larry Ellison, David Friedberg, Jensen Huang, John Martinis, Bob Mumgaard, Lisa Su, and Mark Zuckerberg. Together they represent an extraordinary concentration of market power: the dominant AI chip manufacturer, the leading social AI platform, the cloud infrastructure backbone of the TikTok divestiture deal, the co-founder of the company building the most capable AI assistant on earth, and the semiconductor CEO competing for the next generation of AI training hardware.
PCAST's first policy recommendations are expected within 90 days, arriving as Congress weighs AI legislation and the Commerce Department finalizes semiconductor export rules.
The executives who shape those markets will now formally advise the government writing the rules that govern them. That sentence deserves to sit for a moment before the rest of the analysis begins.
The Full Council: Who's In and What They Control
The council will be co-chaired by White House AI and crypto czar David Sacks and technology adviser Michael Kratsios. The co-chairs define the frame: Sacks, who has been publicly critical of AI regulation and has favored innovation velocity over precautionary oversight, sets the ideological temperature of a council that will advise on the most consequential technology policy decisions of the decade.
The 13 confirmed members and their strategic relevance:
| Member | Role | Strategic Significance |
|---|---|---|
| Jensen Huang | Nvidia CEO | Controls ~80% of AI training chip market |
| Mark Zuckerberg | Meta CEO | Leads Llama open-source AI; largest social AI deployment |
| Larry Ellison | Oracle Executive Chairman | Backbone of TikTok divestiture; Stargate infrastructure partner |
| Sergey Brin | Google Co-Founder | Returned to lead Gemini AI race at Alphabet |
| Lisa Su | AMD CEO | Primary US alternative to Nvidia's AI chip monopoly |
| Marc Andreessen | a16z Co-Founder | Largest AI VC; significant Trump donor and policy influence |
| Safra Catz | Oracle EVP | Enterprise cloud and government contract infrastructure |
| Michael Dell | Dell CEO | AI hardware and enterprise infrastructure |
| Bob Mumgaard | Commonwealth Fusion Systems CEO | Nuclear fusion energy for AI compute power |
| Jacob DeWitte | Oklo CEO | Nuclear power for AI data centers |
| Fred Ehrsam | Coinbase Co-Founder | Crypto infrastructure aligned with Sacks's priorities |
| David Friedberg | Ohalo Genetics CEO | Agricultural biotech and food security |
| John Martinis | UC Santa Barbara Professor | Quantum computing — the panel's sole academic scientist |
Industry experience has taken clear priority over academic independence in this iteration. Only one traditional scientist, quantum computing researcher John Martinis, sits on a panel otherwise dominated by executives running companies with trillions of dollars in combined market capitalization — all of it exposed to the policy areas PCAST is tasked with advising on.
The presence of two nuclear energy executives — Mumgaard and DeWitte — is not incidental. Artificial intelligence has emerged as a major driver of U.S. investment, with companies pledging trillions of dollars in spending over the coming years. The data center power demands that underpin those investments have made nuclear energy a de facto AI infrastructure question, not merely an energy policy question. Having the CEOs of Commonwealth Fusion Systems and Oklo on the AI advisory council signals that the administration views AI compute power and energy generation as a unified policy domain.
The Notable Exclusions: Musk, Altman, and What Their Absence Signals
The invitations that were not extended reveal as much about the administration's political geometry as the appointments themselves.
Elon Musk and Sam Altman were not included in Trump's council. Both exclusions are remarkable for different reasons — and neither is accidental.
Musk ran DOGE, has been arguably Trump's most visible Silicon Valley ally, and controls xAI's Grok, a direct competitor to the AI systems represented on the council. His absence from PCAST may reflect the administration's desire to separate its policy advisory function from its most polarizing political ally — or it may reflect the growing tension between Musk's government role and his competitive interests in AI. Either way, the world's most famous technology supporter of the Trump presidency is not in the room where AI policy will be made.
Altman's exclusion is more pointed. OpenAI is the company that catalyzed the current AI investment supercycle, the organization whose ChatGPT first made AI a mainstream technology, and the firm at the center of the Stargate announcement that Trump celebrated as a signature economic achievement. Excluding Altman while appointing Oracle's Ellison — whose company is a Stargate infrastructure partner — suggests the administration sees OpenAI as a vendor relationship rather than a policy partner. The distinction matters: vendors execute; advisors shape.
The Financial Alignment: Following the Money
Here's what most coverage of PCAST has treated as a footnote but should be the headline.
Google, Meta, and Nvidia each donated one million dollars to Trump's inauguration committee. Marc Andreessen has separately donated to super PACs supporting Trump. Meta, Google, and Huang all contributed to the construction of a new White House ballroom.
It's worth noting that Meta, Google and Huang all chipped in to help pay for the construction of Trump's White House ballroom. Meanwhile, Ellison — whose family has spent much of the last couple of years building a media empire — has a history of support for Trump.
Meta donated $1 million to Trump's 2025 inauguration fund, demonstrating Zuckerberg's renewed relationship with the president, who in return has been less critical of Zuckerberg. Oracle has had a better working relationship with Trump for years because of Ellison's support. Last year, Ellison partnered with Trump to announce a major $500 billion Stargate investment in AI infrastructure.
Huang has regularly met with Trump in a largely successful effort to convince the White House to lift restrictions on the sale of Nvidia's AI chips to China. Andreessen's venture capitalist firm has played a major role in how the White House and congressional Republicans have approached AI policy.
The pattern is not subtle. The companies most visibly aligned with the Trump administration financially — Meta, Nvidia, Oracle, Andreessen Horowitz — now hold formal advisory positions on the panel designing the rules that govern their markets.
The Conflict of Interest Architecture
Meta faces ongoing regulatory proceedings. Nvidia faces scrutiny over its market dominance in AI chips. Google is defending itself in multiple antitrust cases. With PCAST appointments, these companies now hold direct channels to shape rules governing their own markets.
This is the structural tension that PCAST's formation embeds into American AI governance — and it is not without precedent. Under Obama, PCAST included then-Google executive chairman Eric Schmidt and former Microsoft chief research officer Craig Mundie. Biden's panel also included Lisa Su, and ex-Disney CEO Bob Iger served on PCAST during Trump's first term.
What distinguishes this iteration is the concentration of active tech CEOs whose companies collectively face regulatory action from the same government they now advise. Schmidt was Google's chairman when he joined Obama's PCAST — but Google was not simultaneously defending itself in landmark antitrust proceedings with the same administration's Justice Department. The 2026 version of PCAST contains companies in active regulatory crosshairs sitting at the table where regulatory posture is discussed.
Whether the advisory role translates into regulatory protection is an empirical question that will be answered over the next 90 days, when PCAST's first recommendations are due. The structural incentives, however, are legible and uncommented on in the official White House press release.
The China Framing: AI Advisory as Geopolitical Signal
The council is part of a broader effort by the Trump administration to position the United States as a global leader in AI and advanced tech sectors. Its core focus will include AI research, chip development, workforce strategy, and national security. The council's mandate covers several strategic priorities, chief among them advancing U.S. leadership in AI research, both in commercial applications and government deployment.
Trump established PCAST by executive order in January 2025, saying that "as our global competitors race to exploit these technologies, it is a national security imperative for the United States to achieve and maintain unique leadership" in AI and related technologies.
The China competition framing is the political architecture that makes PCAST's formation legible to both political parties. Few politicians will argue against a council whose mandate includes "national security" and "maintaining AI leadership." The framing also creates a useful ambiguity: the same export control decisions that might appear to be regulatory capture when viewed through a domestic competition lens look like national security strategy when viewed through the China lens.
Huang's presence is particularly charged in this context. He has regularly met with Trump in a largely successful effort to convince the White House to lift restrictions on the sale of Nvidia's AI chips to China. Huang now sits on the panel that will advise on semiconductor export policy — the specific policy area his direct lobbying has been designed to influence. The advisory role does not change the substance of his position. It changes the formality of the channel through which that position reaches the Oval Office.
The PCAST Mandate: What It Will Actually Do
A press release from the Office of Science and Technology Policy said PCAST "brings together the Nation's foremost luminaries in science and technology to advise the President and provide recommendations on strengthening American leadership in science and technology." The council will focus on topics "related to the opportunities and challenges that emerging technologies present to the American workforce, and ensuring all Americans thrive in the Golden Age of Innovation."
In practical terms, the council's work over the next 90 days will intersect with several live policy decisions: the Commerce Department's finalization of semiconductor export rules, competing AI legislation in Congress, the AI Safety Institute's operational mandate, and the administration's Stargate implementation timeline.
With co-chairs who have publicly favored innovation velocity over precautionary regulation, those recommendations could shape the terms of federal AI oversight at a pivotal moment for the industry. Sacks in particular has been unambiguous: he criticized the AI Overwatch Act in January 2026, claiming it would undermine Trump's authority over AI chip exports. The co-chair of the advisory panel responsible for AI policy recommendations has already staked out a position on AI chip export legislation.
Practical Takeaways: What This Means for the AI Industry
- Semiconductor export policy is the highest-stakes near-term decision area — with Huang and Su both on the council, the advisory positions on chip export rules will reflect the interests of the two dominant US AI chip manufacturers simultaneously
- Regulatory proceedings against Meta, Google, and Nvidia will continue through the DOJ and FTC, but the executives now have a formal government advisory channel that did not exist before March 25
- AI safety legislation faces a more industry-favorable environment with Sacks co-chairing the panel that will shape the administration's position
- Nuclear energy investment for AI data centers is signaled as a policy priority by the inclusion of two nuclear startup CEOs — expect executive orders or federal incentives supporting nuclear power for compute within the council's first recommendation cycle
- Open-source AI policy will likely reflect Zuckerberg's strong pro-open-source position, which aligns with Meta's Llama strategy and creates tension with the national security arguments for keeping frontier models proprietary
The Architecture of Captured Governance
There is a version of PCAST that functions exactly as advertised: a panel of people who understand AI technology better than most government officials, providing technical expertise that improves federal decision-making. This version is real and the expertise on the council is genuine. Jensen Huang understands semiconductor manufacturing chains better than any congressional staffer. Lisa Su understands AI chip design with a depth that no career civil servant can match. Mark Zuckerberg runs the largest open-source AI research program in the world.
But expertise and interest are not the same thing — and the PCAST roster conflates them in ways that the official framing does not acknowledge.
Shoshana Zuboff would recognize the structure immediately: the regulatory apparatus is being staffed by the entities whose behavior it is designed to regulate, under the rubric of "expertise." The companies with the most to gain from favorable AI chip export policy, permissive AI content regulation, and light-touch antitrust enforcement now hold formal advisory positions on each of those exact policy domains.
History suggests that advisory panels with this composition tend to produce recommendations that prioritize industry velocity over regulatory caution, market incumbency over open competition, and national security framing over worker protection. Whether that serves the American public interest depends almost entirely on one's prior beliefs about whether accelerating AI development is the primary obligation of federal AI policy — or whether governing it responsibly is.
PCAST cannot answer that question. Only the political process can. But the 13 people now formally advising on the answer have a combined market capitalization exceeding $5 trillion — and every one of them has a financial interest in how it resolves.
The council that will shape American AI governance for the next decade was announced on March 25, 2026. The question that announcement does not answer — and that no White House press release is designed to answer — is this: when the companies being regulated are the ones advising on regulation, whose definition of the "Golden Age of Innovation" are we actually building?
My Take:
The Boardroom is the New Oval Office "Trump’s PCAST isn't just an advisory board; it’s a 'War Cabinet' for the AI age. By putting Jensen Huang (Nvidia) and Mark Zuckerberg (Meta) at the table while excluding Sam Altman (OpenAI), the administration is signaling a shift from 'AI Safety' to 'AI Dominance.' My advice to the YousfiTech audience: Don't just watch the appointments, watch the Chip Export Rules. Jensen Huang has been lobbying to lift China restrictions for years—now, he’s the one writing the advice on how to do it. We are witnessing the ultimate convergence of corporate interest and national security. In 2026, the 'Golden Age of Innovation' is being designed by the very people who will profit the most from it."
🔗 Internal Linking Suggestions for YousfiTech AI
- "Nvidia's AI Chip Monopoly: How Jensen Huang Built an $3 Trillion Company — and the Risks Ahead" — deep profile of Nvidia's market dominance, the export control battles Huang has been fighting, and what PCAST membership means for the company's regulatory environment
- "Big Tech's Political Pivot: How Meta, Google, and Nvidia Bet on Trump's Second Term" — analysis of Silicon Valley's strategic alignment with the Trump administration since 2024, the inauguration donations, the dinner diplomacy, and what each company got in return
0 Comments