Musk and Zuckerberg’s G20 message: build more AI data centers, regulate less

The message from Elon Musk and Mark Zuckerberg at the G20 Innovation Ministerial was straightforward: governments should help AI scale faster by enabling more data centers, securing abundant electricity, supporting startups and keeping regulation light. The September 1–2 meeting in Chapel Hill, North Carolina, turned AI infrastructure into an economic and geopolitical issue—not just a technology-policy debate. 17

That pro-buildout argument met a growing public backlash. Opponents of new data centers have raised concerns about higher electricity bills and persistent noise, creating a direct conflict between national ambitions for AI leadership and the local costs of hosting the infrastructure that makes it possible.

What Musk and Zuckerberg asked governments to do

Expand data centers and the power needed to run them

Musk and Zuckerberg argued that the AI race depends on physical infrastructure. That means building more data centers and increasing the electricity supply available to them, rather than treating compute capacity as a problem that can be solved through chips alone.

Musk specifically urged countries to develop new energy sources outside China. His strategic concern was that the country able to add generating capacity most quickly could gain an advantage in AI deployment. Reports from the meeting also attributed to him a warning of a significant power shortfall as early as the following year.

Prefer light-touch AI rules

Musk criticized European-style regulation as a brake on innovation and advocated relatively limited rules and greater reliance on self-regulation. The broader U.S. position at the meeting favored avoiding entirely new AI regulations where existing laws could address the issue. 17

That approach was presented as a way to keep emerging technologies deployable and competitive. It was not a consensus finding that regulation has no role: the meeting took place amid an international debate over how to balance AI’s economic potential with its risks. 12

Direct support toward startups and new entrants

The executives also linked AI competitiveness to entrepreneurship. Zuckerberg said AI could dramatically reduce the cost of starting a business, while the policy argument at the meeting favored ensuring that support for innovation reaches startups rather than concentrating it only among established technology companies. 913

Musk’s forecasts about AI’s economic impact

Musk used the meeting to make several far-reaching predictions. He described AI as capable of making goods and services dramatically cheaper and suggested that it could eventually handle essentially all digital-sector work, including software development. He also said AI software could reach “Stockfish level” in roughly 12 to 18 months—an analogy to the strength of a leading chess engine—and argued that humans would no longer be able to compete with AI in software development at that point.

These statements should be read as Musk’s forecasts, not as conclusions adopted by the G20 or independently validated projections. The available reporting describes the claims and the policy debate around them, but it does not establish that the timelines will be met.

Musk also predicted a rapid expansion of humanoid robots and portrayed robotics as a potentially transformative source of productivity. The supplied reporting supports the scale and ambition of that forecast, but not a verified timetable or outcome.

Zuckerberg’s workforce warning

Zuckerberg’s argument added a practical constraint to the infrastructure push: data centers require people to build and maintain them. Reporting from the meeting highlighted shortages of skilled workers needed for data-center construction, including trades such as electricians, construction workers and plumbers.

That matters because faster AI deployment depends on more than investment, permits and hardware. If the industry cannot find enough qualified workers, labor availability could slow construction even when companies have financing and access to power.

Why the local protests mattered

The meeting’s infrastructure message collided with opposition in communities where new data centers are proposed. Critics have focused on the effect of large facilities on electricity demand, household bills and neighborhood noise.

Musk defended data centers as necessary infrastructure for the AI buildout. President Donald Trump also criticized the backlash, warning that opposition could benefit China rather than treating community objections as a reason to slow development.

This created the central tension of the meeting: the benefits of AI infrastructure are described in national terms—economic growth, technological leadership and competition with China—while many of the immediate burdens are local. Residents near proposed facilities may experience the effects through land use, noise, grid investment or utility costs before any broader economic gains become visible.

Where the meeting fits in the G20 calendar

The Chapel Hill event was a technology-focused G20 ministerial co-hosted by the U.S. Department of Commerce and the White House Office of Science and Technology Policy. It brought together commerce and technology ministers representing more than 20 countries or regions. 12

It was part of the run-up to the G20 leaders’ summit scheduled for December 14–15 in Miami. The ministerial therefore offered the U.S. administration and technology executives a venue to promote a pro-innovation, light-regulation approach before the leaders’ meeting.

The unresolved policy trade-off

Musk and Zuckerberg presented a single AI-competitiveness package: lighter regulation, more startup support, more data centers, more electricity and a larger skilled-trades workforce. Their case is that slowing any part of that chain could reduce economic gains and weaken the United States and its partners in competition with China.

The protests exposed what the package leaves unresolved. Governments still have to decide how power costs, noise, land use, grid upgrades and other infrastructure pressures should be shared. The Chapel Hill debate was therefore not simply about whether to build AI capacity. It was also about who receives the benefits, who carries the risks and what safeguards should accompany the speed that Musk and Zuckerberg want.