Zuckerberg and Amodei Head to Trump Lunch as META Falls 4.8%

Meta's Mark Zuckerberg and Anthropic's Dario Amodei are among technology executives scheduled to meet President Trump on Tuesday, with Meta shares (META) trading at 715.43, down 4.82% on the day as of 19:54 GMT on Sept. 28, 2026.

Mark Zuckerberg and Dario Amodei are among the technology executives scheduled to sit down with President Donald Trump on Tuesday, according to CNBC Technology. The meeting lands in the middle of an unusually public argument inside the industry about how fast frontier artificial intelligence models should be built and who, if anyone, should be allowed to slow them down.

That argument is no longer confined to research labs. Amodei, Elon Musk and OpenAI chief executive Sam Altman have all called in recent weeks for pacing the development of frontier models — the largest, most capable systems, the ones whose behavior cannot be fully predicted before training finishes. When three of the people best positioned to accelerate say the pace itself is the problem, the policy question stops being theoretical.

Two very different asks walking into the same room

Zuckerberg and Amodei arrive with close to opposite incentives, which is what makes the guest list worth reading carefully.

Meta has built its AI position on scale and distribution: models pushed into products used by billions, and an open-weights posture that puts capability into the hands of outside developers. Anything that imposes pre-release licensing, capability thresholds or liability for downstream misuse falls heaviest on that model. Meta's interest in Washington is generally in keeping federal rules light, preemptive of a patchwork of state laws, and workable for systems that ship to consumers rather than sitting behind enterprise contracts.

Anthropic, which Amodei runs, has made safety framing central to its identity and its enterprise sales pitch. A company that already runs internal evaluation regimes and staged deployment gates has less to lose from rules that require them — and something to gain if competitors must build the same apparatus. That is not cynicism; it is simply how compliance costs distribute. Firms already carrying the cost prefer it become universal.

Musk and Altman are not confirmed attendees in the reporting, but their public positions frame the conversation. Both have argued for pacing frontier development while continuing to build at the frontier. The tension in that stance — asking for brakes on a road you are still racing down — is precisely what a president has to adjudicate if any federal framework is going to emerge.

Meta shares slide while the broader tape sags

The meeting news arrived on a weak session for equities. As of the last trade at 19:54 GMT on Monday, Sept. 28, 2026, META changed hands at 715.43, down 4.82% from its previous close of 751.66 — a fall of 36.23 points on the day, on illustrative arithmetic from those two prices. The stock traded between 715.00 and 750.58, and was sitting at the bottom of that range, which is the shape of a session where sellers had the last word rather than one that stabilized after an early drop.

The benchmarks were softer but nothing like as soft. The S&P 500 tracker (SPY) was at $765.19, down 0.80% from a prior close of $771.35. The Nasdaq 100 tracker (QQQ) sat at $736.13, off 1.12% against $744.50, with a day range of $731.63 to $741.42. The Dow 30 tracker (DIA) traded at $514.01, down 0.67%. Meta's decline was therefore several times the size of the tech benchmark's, which points to company- or sector-specific pressure rather than a simple beta move.

No single cause should be assigned to a one-day move on the strength of a meeting invitation. But the setup matters for how investors read Tuesday: a large-cap AI name giving up ground into a White House session where the policy direction for frontier models is on the table.

Why the regulatory question reaches the share price

AI policy has been treated by markets mostly as a narrative input — something that affects sentiment around the theme rather than the cash flows of specific companies. That gap is closing. Three mechanisms turn a rulebook into an earnings input.

• Time to market. Pre-deployment evaluation requirements add months between a finished model and a shipped product. For companies monetizing through consumer features, delay is revenue deferred.
• Fixed compliance cost. Safety testing, red-teaming and documentation are largely fixed expenses. They are trivial against a hyperscaler's budget and meaningful against a startup's, which concentrates the market rather than opening it.
• Open weights. Any rule that attaches responsibility to how a released model is later used strikes directly at the open-release strategy. That is the single policy lever with the most asymmetric effect across the named companies.

Capital spending is the other channel. The AI build-out — data centers, power contracts, accelerator purchases — is underwritten by an assumption that frontier capability keeps improving and keeps being deployable. A federal posture that formally paces frontier development would not stop that spending, but it would change the timing assumptions underneath it, and timing is most of what a discounted cash flow model is.

What to watch after Tuesday

The substantive tells will not be in the readout photos. Watch for whether any federal framework is described as preempting state AI laws, which is the outcome large platforms most want and which would remove a real source of legal uncertainty. Watch whether compute thresholds reappear as the trigger for obligations, since that draws a bright line between who is regulated and who is not. And watch whether export controls and chip access are folded into the same conversation, because that is where AI policy stops being a domestic matter and starts moving semiconductor revenue.

For shareholders, the near-term question is narrower: whether Meta's Monday weakness is a one-session repricing or the start of investors discounting policy risk they had previously ignored. Tuesday's meeting will not answer that, but it will show which version of the safety argument has the president's ear.

Frequently asked questions

Who is meeting President Trump on Tuesday?

Meta chief executive Mark Zuckerberg and Anthropic chief executive Dario Amodei are among the technology executives set to meet President Donald Trump on Tuesday, as reported by CNBC Technology. The full attendee list was not detailed in that report, and neither Elon Musk nor OpenAI's Sam Altman was confirmed as present despite both being central to the current debate.

What is the frontier-model debate about?

Frontier models are the largest and most capable AI systems, whose behavior cannot be fully predicted before training completes. Amodei, Elon Musk and OpenAI chief executive Sam Altman have all called in recent weeks for pacing their development. The dispute is over whether that pacing should be voluntary, industry-led, or written into federal rules with enforcement behind it.

How did Meta stock trade on Monday?

META traded at 715.43 as of the last trade at 19:54 GMT on Sept. 28, 2026, down 4.82% from a previous close of 751.66. The day range was 715.00 to 750.58, meaning the stock was sitting at the low end of its session range rather than recovering from an earlier decline.

Was the whole market down that day?

Yes, but far less. The S&P 500 tracker SPY was at $765.19, down 0.80%. The Nasdaq 100 tracker QQQ traded at $736.13, off 1.12%. The Dow 30 tracker DIA was at $514.01, down 0.67%. Meta's decline was several times larger than the tech benchmark's move on the same session.

Why would AI regulation affect company earnings?

Three channels matter: pre-deployment testing requirements delay product launches and defer revenue; compliance costs are largely fixed, so they weigh more on smaller entrants than on hyperscalers; and any liability attached to how released models are later used strikes directly at open-weights release strategies. Capital spending assumptions for data centers also depend on deployment timing.

What should investors watch after the meeting?

Whether any proposed federal framework preempts state AI laws, which large platforms favor for removing legal uncertainty. Whether compute thresholds return as the trigger for regulatory obligations, which would define who is covered. And whether chip export controls are folded into the same policy discussion, since that would move semiconductor revenue as well as software timelines.

Sources

• Zuckerberg, Amodei among tech executives set to meet Trump Tuesday — CNBC Technology