A figure never heard in a recent institutional context: Mark Zuckerberg, during a dinner at the White House on September 4, 2025 Euronews, spoke of a plan for $600 billion in investments in the United States by 2028.
As of September 5, 2025, however, there are no official communications or memorandums published in Meta’s newsroom Meta Newsroom, a circumstance that reinforces the exploratory nature of the public mention.
The announcement, broadcast live by broadcasters and video platforms, sparked an immediate debate on public priorities, private responsibilities, and the ability of infrastructure to support the evolution of artificial intelligence.
In this context, questions about governance and execution timelines multiplied immediately.
According to the data collected by our editorial team, based on video recordings of the event and checks on official channels, no formal contractual documents linked to this amount had been filed as of September 5, 2025.
Industry analysts consulted note that a 600 billion plan, if distributed until 2028, would average about 150 billion dollars per year, typically disbursed in tranches subject to permits, authorizations, and regulatory obligations. These preliminary estimates are intended to guide subsequent verifications, not to define binding commitments.
The Key Moment at the Dinner with Trump
At one point during the event – during which President Trump hosted the leaders of the tech sector – videos and press reports indicate that Zuckerberg stated, quite informally, that Meta expected “probably, I don’t know, at least 600 billion” dollars in the USA by 2028.
The comment, despite being a statement not yet anchored to a contractualized program Axios, has captured the attention of politics, the market, and regulators. That said, the tone appeared exploratory and non-binding, a signal that leaves room for further clarifications.
Who was there and what was discussed
Approximately 33 leading figures from the technology sector attended the evening. Among those mentioned by U.S. media were Mark Zuckerberg, Sam Altman, Bill Gates, and Tim Cook; notably, contrary to some initial rumors, Elon Musk did not attend the event Sky TG24.
The discussion table focused on international competitiveness, the semiconductors supply chain, incentives for data centers, and domestic production, with a focus on generative AI. However, the talks also addressed practical issues such as authorization times and capital allocation.
- Key participants: top managers of Big Tech and founders, in a rare moment of cross-sectional discussion.
- Stated objective: illustrate investment plans and discuss industrial policies and rules for AI.
- Coverage: video footage and reports, including Axios and Sky TG24.
What does “600 billion” mean: scope and nature of the expenses
The time reference indicated is “until 2028.” It is not yet clear whether this involves multi-year CapEx, recurring operating expenses, public-private partnerships, or a combination of these elements.
Currently, the commitment appears as a verbal statement rather than a contractualized program with definitive timelines and locations. Indeed, the issue of annual allocation and phase sequencing remains open.
Where the funds might end up
- Construction and expansion of high-efficiency data centers.
- Purchases of AI accelerators and computing infrastructure.
- Connectivity networks, enhancement of electrical transmission and energy supply.
- Applied research, model security, and technical training.
- Support for the semiconductor ecosystem and critical supplies.
Regulatory Reactions and Cautions
The announcement has sparked a mix of caution and interest. Analysts and institutional observers are calling for greater clarity on timelines, site locations, employment impacts, and tax conditions.
Some critical positions suggest linking any tariff advantages to verifiable obligations in terms of sustainability, security, and competition. It should be noted that the most recurring request concerns standardized metrics and accessible monitoring criteria.
Potential Impact on AI, Data Centers, and Chips
A plan of this magnitude could accelerate the adoption of AI, unlock collateral investments in energy and the chip supply chain, and strengthen the US position in the global landscape, while some aspects of transparency such as energy costs, water consumption of data centers, security standards, and model governance remain to be clarified.
That said, the actual scope will depend on the synchronization between industrial policy and the execution capacity of the parties involved.
What We Know So Far
- Fact: during the dinner on September 4, 2025, a plan of 600 billion dollars by 2028 was mentioned, although the wording remained informal.
- Under discussion: the nature of the amounts (CapEx vs Opex), the extent of public incentives, the actual number of sites, and geographical priorities.
- To verify: the publication of official documents, timelines, contracts, employment clauses, and environmental provisions.
Indicators to Monitor
- The publication of memorandum or framework agreements with federal and state governments.
- The groundbreaking for data centers, with building and energy permits filed.
- The orders and deliveries of GPU/accelerators reported by listed suppliers.
- The impact on local employment and training programs, through updated datasets of qualified profiles.
- Measurable commitments on energy efficiency, water consumption, and the use of renewable sources.
What to Check Immediately
- Official announcements: updates from Meta’s newsroom and the Briefing Room page of the White House for full text and video confirmations.
- Publication of contracts: details regarding state/municipal incentives and the related attachments through public registers.
- Employment impact: estimates and actuals on direct and induced jobs in the affected areas.
In summary
The announcement of a possible 600 billion plan that emerged during a dinner at the White House represents a signal of industrial and technological ambition without recent precedent.
The credibility of the project will depend on transparency and the potential publication of official documents and verifiable milestones, essential elements for measuring the real effect on investments in AI, infrastructure, and US competitiveness. Ultimately, the gap between promise and realization will be measured by the formal steps that follow.