This article contains affiliate links. We may earn a commission at no extra cost to you if you sign up through our links.
Trump's AI Dream Team: Zuckerberg, Huang, Brin In — Musk and Altman Out
Trump's first 13 appointments to the White House AI council include the CEOs of Nvidia, Meta, and AMD — but the two most famous names in AI were conspicuously missing.
On March 25, 2026, President Trump named 13 tech executives to the President's Council of Advisors on Science and Technology (PCAST). In: Mark Zuckerberg (Meta), Jensen Huang (Nvidia), Larry Ellison (Oracle), Sergey Brin (Google co-founder), Lisa Su (AMD), Marc Andreessen (a16z), and nuclear fusion CEOs. Out: Elon Musk and Sam Altman. The council is co-chaired by AI czar David Sacks and will advise on chip exports, data center policy, and US AI competitiveness vs. China.
The Full Council Roster
On March 25, 2026, the White House announced the first 13 members of PCAST — the President's Council of Advisors on Science and Technology. The body was created by executive order in January 2026 to advise on artificial intelligence, emerging technologies, and national innovation strategy. It is co-chaired by David Sacks, the White House AI and crypto czar, and Michael Kratsios, director of the Office of Science and Technology Policy.
| Name | Company / Role | What They Represent |
|---|---|---|
| Mark Zuckerberg | CEO, Meta Platforms | Open-source AI models, social AI platforms |
| Jensen Huang | CEO, Nvidia | AI chip dominance, GPU compute infrastructure |
| Larry Ellison | Executive Chairman, Oracle | Enterprise cloud, AI data centers |
| Sergey Brin | Co-founder, Google | Search AI, foundation model research |
| Lisa Su | CEO, AMD | Chip competition, alternative GPU supply chain |
| Michael Dell | CEO, Dell Technologies | Enterprise hardware, AI server market |
| Marc Andreessen | Partner, a16z | Venture capital, AI regulatory philosophy |
| Safra Catz | CEO, Oracle | Enterprise data, government cloud contracts |
| Jacob DeWitte | CEO, Oklo Inc. | Nuclear fission power for AI data centers |
| Bob Mumgaard | CEO, Commonwealth Fusion | Nuclear fusion energy for AI infrastructure |
| Fred Ehrsam | Co-founder, Coinbase | Crypto, decentralized AI finance |
| David Friedberg | Founder, The Climate Corporation | AI in agriculture and climate |
| John Martinis | Google physicist | Quantum computing research |
The Two Biggest Names Not on the List
The most significant aspect of the March 25 announcement was not who was included — it was who was not. Two names that define the global AI conversation were absent from the initial roster: Elon Musk and Sam Altman.
Musk's exclusion is notable given his role in founding DOGE and his previous access to Trump administration decision-making. Altman's exclusion is notable because OpenAI is the company most associated with the current AI wave in the public consciousness. Neither received a call.
Musk and Altman are currently engaged in active federal litigation against each other. Musk sued OpenAI over its for-profit conversion; Altman counter-sued alleging monopolization attempts and interference. Placing both on the same 13-person advisory council would create an unmanageable conflict of interest and potential legal liability for the administration. The cleanest solution was to exclude both. The White House has not officially confirmed this reasoning.
A Hardware-First Strategy — By Design
The composition of the council is not accidental. It reflects a deliberate "hardware-first" philosophy about what matters most in the AI race with China. The council is dominated by executives who control the physical layer of AI: Nvidia and AMD own chip design, Oracle and Dell own data center infrastructure, and Oklo and Commonwealth Fusion represent the energy supply for future compute.
The policy implications are predictable from the roster. Nvidia and AMD will push for loosened chip export controls — currently restricting sales of H100s and A100s to Chinese buyers. Oracle and Dell will push for streamlined federal permitting for hyperscale data centers. The nuclear CEOs signal a strategic focus on solving AI's energy constraint: next-generation data centers require power that current grids cannot reliably deliver.
What the council notably does not include: any academic AI researcher, any AI safety specialist, any ethicist, or any representative of the labor market disruption that AI is creating. Every member either sells AI infrastructure or represents capital that profits from its continued growth.
The presence of Jensen Huang (Nvidia) and Lisa Su (AMD) on the council is the most commercially significant signal in the announcement. Both companies are directly constrained by current export restrictions that prevent selling their most powerful chips to China. Huang has publicly argued that these restrictions hurt US companies more than they hurt China, which develops domestic alternatives. With two chip CEOs advising directly on policy, a loosening of export restrictions is more likely in 2026 than it has been at any point since the restrictions were introduced.
The Conflict of Interest Problem
Critics have noted that every major appointee represents a company with a direct financial interest in specific government policy outcomes. Nvidia needs export control relief. Oracle needs streamlined data center permitting and continued government cloud contracts. AMD needs the same chip-export flexibility as Nvidia. Marc Andreessen's portfolio companies need permissive AI regulation. Even the nuclear energy executives are seeking regulatory streamlining and federal energy contracts.
The pattern is not unique to this administration — industry advisory councils have always skewed toward industry — but the concentration in the AI sector is notable precisely because AI policy decisions made in 2026 will shape market structure for decades. An advisory body composed entirely of companies that need something from the government is a different proposition than an advisory body that includes independent voices.
PCAST vs. Prior AI Policy Bodies
| Body | Year | Industry Reps | Safety Voices | Academic Reps |
|---|---|---|---|---|
| Trump PCAST 2026 | 2026 | 13/13 (100%) | 0 | 1 (Martinis) |
| Biden AI EO Task Force | 2023 | ~60% | 4 (NIST, CISA) | ~5 |
| EU AI Office Advisors | 2024 | ~50% | ~8 | ~10 |
| UK AI Safety Institute | 2023 | ~40% | ~10 | ~12 |
What It Means for the AI Industry
The practical consequences of the PCAST's composition will play out over months, not weeks. Advisory councils advise — they do not legislate. But the access these executives now have to the White House on AI regulation creates a direct channel from boardroom to policy, at a moment when the US is actively rewriting its approach to AI governance.
For Anthropic and OpenAI — both absent from the council — this represents a structural disadvantage in policy conversations. Regulatory outcomes that favor hardware and infrastructure providers over model-layer companies could slow Anthropic's enterprise penetration or constrain OpenAI's access to compute. This is one more reason both companies are now racing to go public: capital independence matters when your main competitors have direct advisory access to the administration.
Frequently Asked Questions
Comments are coming soon.