Analysis
Nearly 20 technology CEOs met privately with President Trump at a White House lunch this week, and The Information's look at who actually had power in the room argues the guest list itself is the story -- a rough map of who Washington currently treats as AI's decision-makers. Attendees included Meta's Mark Zuckerberg, Elon Musk, Anthropic's Dario Amodei, Google's Sundar Pichai, Nvidia's Jensen Huang, Palantir's Alex Karp and Shyam Sankar, Amazon's Jeff Bezos and OpenAI's Greg Brockman, according to Axios's full attendee list.
Self-Regulation Over Statute
Trump emerged from the summit calling for "tremendous self-regulation" by AI companies rather than committing to new federal rules, with attendees reportedly signing onto a voluntary "constitution" for policing themselves. That is notably softer than the direction Congress has been moving on a separate track -- the same week, the Senate held its first hearing on rogue AI agents and a bill extending hacking-liability law to AI operators began moving through committee, a parallel push Pulse covered in its look at the AI agent liability fight.
“- Alex Karp and Shyam Sankar (Palantir): representing the defense-and-government side of AI, a different constituency than the consumer labs in the room.”
The attendee list maps onto distinct power bases:
- Dario Amodei (Anthropic): attended despite Anthropic having just paid a record $1.5 billion copyright settlement -- a seat at the table that survived a costly legal loss.
- Jensen Huang (Nvidia): present as the chip supplier whose products underpin every other attendee's AI buildout, giving him leverage regardless of which lab wins any given product cycle.
- Alex Karp and Shyam Sankar (Palantir): representing the defense-and-government side of AI, a different constituency than the consumer labs in the room.
- Jeff Bezos (Amazon) and Greg Brockman (OpenAI): both companies are simultaneously competitors and partners -- Amazon holds a major investor-and-customer relationship with Anthropic through AWS, while OpenAI just closed its pre-IPO round with Nvidia and SoftBank topping it up.
Trump is also reportedly preparing to name Director of National Intelligence Jay Clayton as a new AI czar, per the Washington Post's reporting. Choosing a personnel appointment over new legislation is consistent with the self-regulation framing from the summit itself, and keeps rulemaking inside the executive branch rather than Congress.
For founders and investors trying to read Washington's AI posture, the summit guest list is a more useful signal than any public statement -- it shows which companies currently have direct access to the administration, and by extension, early warning on any coming executive action. Startups without a seat at tables like this one are effectively relying on their larger competitors and partners to represent their regulatory interests.
Forbes noted that women were conspicuously absent from the summit's attendee list despite leading much of the safety and policy work inside these same companies, a gap that undercuts any claim the room represented a full picture of AI's decision-makers rather than just its most visible CEOs. A dinner photo-op and a voluntary "constitution" also commit these companies to nothing enforceable; self-regulation pledges from past tech summits have a mixed record of actually changing company behavior once the cameras leave.
The self-regulation framing buys the industry time, but it doesn't make the agent-liability bill moving through Congress go away -- if anything, a voluntary pledge with no enforcement mechanism gives legislators a cleaner argument for why statute is still necessary.