U.S. President Donald Trump and the largest AI companies have effectively aligned their economic and political interests around the development of artificial intelligence. At the heart of this alliance are four major bets: on the economic payoff from massive infrastructure investment, voluntary regulation, the ability to safely develop increasingly powerful models, and a political partnership between the White House and the tech industry, Axios.com reports.

Whether these expectations are fulfilled will determine more than just the fortunes of individual companies. Success could mean faster productivity growth and a new technological leap for the United States, while failure could lead to major financial losses, labor market disruption, and fresh demands for regulation.

Trillions of dollars pinned on a new industrial leap

The first bet concerns the economics of AI. Companies and investors are pouring enormous sums into accelerators, data centers, energy, and other infrastructure, betting that artificial intelligence will become a new engine of productivity.

By some estimates, total spending on AI infrastructure could reach $10.3 trillion by 2032. At the same time, the return on these investments remains debatable. One analysis cited by Axios found that since 2024, the largest cloud companies have spent nearly $1 trillion more on AI than the technology has generated in revenue for them.

For the Trump administration and tech companies, the wager is that today’s spending will eventually turn into a new cycle of economic growth — from automating existing work to creating entirely new industries and services.

The downside is the risk of large-scale overinvestment. If the expected productivity growth fails to materialize, huge investments in computing infrastructure could prove far less profitable than current estimates assume.

Washington is betting on self-regulation

The second bet concerns regulation. Recently, Trump and representatives of the largest AI companies signed an agreement based on voluntary commitments, internal model evaluations, and independent auditing rather than new federal requirements.

This approach allows companies to develop technologies more quickly and set their own testing and oversight procedures. For supporters of this model, international competition also matters: a lighter regulatory environment is supposed to help the United States maintain an edge over China.

But the system raises a fundamental question: what happens if voluntary measures prove insufficient? In that case, Washington would have to tighten oversight only after the consequences of inadequate safety had already become obvious.

The more powerful the models, the harder the safety challenge

The third bet is that AI can become significantly more powerful without turning into a system its creators are unable to control.

Warnings about potential risks are coming in part from artificial intelligence developers themselves. In particular, documents from Anthropic related to the company’s stock market debut consider scenarios in which sufficiently advanced AI could pose a threat to humanity’s survival.

At the same time, the industry sees growing model capabilities as the source of the most significant potential benefits — from accelerating scientific research to developing new medicines and solving complex engineering problems.

In this way, the industry is trying to strike a balance between two directions. On the one hand, models need to become more autonomous and efficient. On the other, control and safety systems must keep pace with the growth of their capabilities.

The technology alliance is becoming a political one

Finally, there is a political bet. Trump has made the development of artificial intelligence one of the key elements of his technology policy and links it to American leadership in competition with China. AI companies, for their part, are interested in a political environment that allows them to build infrastructure and develop models with a minimal number of restrictions.

As a result, the interests of the White House and the tech industry are becoming ever more intertwined. For the Trump administration, AI’s success could become part of a broader economic and technological program. For companies, government policy directly affects energy availability, data center construction, regulation, and the terms of competition.

But this interdependence cuts both ways. If AI development leads to serious economic losses, a large-scale reshaping of the labor market, or major safety incidents, the consequences will affect more than just individual companies.

The public still does not share the industry’s confidence

At the same time, the AI industry still has a trust problem. According to a new Quinnipiac poll, 74% of Americans said they do not trust executives at AI companies. Only 34% believe that artificial intelligence will bring more benefit than harm to their everyday lives.

This creates an important contrast with the scale of the bets Washington and tech companies are making today. Infrastructure investments are already measured in trillions of dollars, yet public confidence that these investments will lead to a noticeable improvement in people’s lives remains significantly lower.

In essence, the United States is now simultaneously conducting the biggest technological experiment in decades and trying to build a political and economic system around it. The outcome will depend not only on how quickly the models continue to advance, but also on whether AI companies can turn computing power into real productivity, ensure system safety, and convince the public of the value of this technology.