A spate of major investments by large tech firms in U.S. artificial-intelligence (AI) companies should be viewed as a sign of vibrant competition that drives innovation. Antitrust intervention to limit such investments would be inappropriate. A cautious approach to antitrust, combined with deregulation, could be the ticket to ensuring continued American leadership in AI.
AI Firms: Major Investments and Dynamic Competition
As a 2025 Forbes analysis explained:
Artificial intelligence companies (whether apps or model makers) typically rely on expensive silicon chips and energy-intensive data centers for computing power to train and run their models and applications.
That fact, coupled with businesses’ realization by that AI offers enormous potential for productivity growth, has created incentives for tech firms to invest heavily in AI enterprises.
Nvidia’s recently announced $100 billion investment in OpenAI is the latest in a series of tech-company alliances with AI firms. OpenAI, Oracle, and SoftBank very recently revealed five massive new AI-infrastructure sites under Project Stargate, bringing their total planned capacity to 7 gigawatts and nearly $400 billion in investment. Other AI firms have obtained major third-party investments as well.
No single firm has become dominant in the AI sector. Indeed, as Computer & Communications Industry Association (CCIA) Chief Economist Trevor Wagener recently explained, the sector is very competitive, characterized by intense rivalry across multiple layers, from hardware to applications. While tech giants dominate, the market features a rich ecosystem of challengers, including startups and open-source projects.
The Layered ‘AI Stack’
The AI sector isn’t monolithic, but rather a set of interconnected technologies. Competition is fierce at every layer of this stack, including:
- Hardware and Computing Infrastructure: Major players like Google, Amazon, Microsoft, and NVIDIA are aggressively developing specialized AI chips to enhance training and performance.
- Foundation Models: There is fierce competition among developers of large language models (LLMs) and other advanced models. The performance gap among the leading models is rapidly shrinking, indicating a crowded and competitive frontier.
- Applications and Services: Many firms build specialized AI applications and tools for end-users. The diversity of players and business models at this layer creates a rich competitive mix.
- Cloud Access: As Wagener described, “top-tier AI startups like Anthropic, Cohere, and Character.ai have struck deals with different cloud providers (some use AWS, others Google Cloud, etc.), ensuring that no single cloud company ‘locks up’ all promising AI firms?. In fact, cloud vendors often offer special incentives, optimized hardware, and dedicated AI services to attract AI developers onto their platforms, which indicates how fiercely they compete to be the go-to deployment venue.”
- Massive Investment: Private and government investment in AI is at an all-time high globally, fueling rapid innovation. In 2024, private U.S. investment and global investment in generative AI both saw significant growth.
- Falling Costs and Increased Accessibility: Advancements in hardware efficiency and open-source models are lowering the barriers to entry for advanced AI. The inference cost for models performing at the level of GPT-3.5 dropped more than 280-fold between late 2022 and late 2024, making advanced AI more accessible.
- Rapid Innovation Cycles: Competition is driving a “virtuous cycle” of innovation where breakthroughs by one firm quickly become the baseline for others. The constant need to evolve prevents any single player from achieving lock-in dominance.
Wagoner sums it up:
AI is proving to be a disruptive force within tech ecosystems, breaking open markets and spawning new entrants rather than entrenching incumbents?. The evidence is clear that competition in AI applications is vibrant and delivering rapid innovation and expanding consumer choice.
In other words, far from threatening competition, AI alliances promote enhanced dynamic competition.
The Specter of Antitrust
Despite all these signs of effective competition, U.S. and foreign antitrust enforcers nevertheless have shown a strong interest in potentially challenging AI-ecosystem arrangements. Antitrust and technology expert Jonathan Barnett of the University of Southern California describes the spate of aggressive investigations into AI markets (including those involving agreements between large tech platforms and smaller AI providers) by trustbusters from the United States, the European Union, and multiple other jurisdictions.
As Barnett emphasizes, antitrust agencies’ focus on AI arrangements “reflects the ascendance of a preemptive approach toward antitrust enforcement in digital markets” that favors early intervention lest those markets “tip” in an anticompetitive direction. Digital markets’ scale economies and network effects are thought to favor the eventual emergence of only one or a limited number of dominant firms. (Antitrust proponents typically fail to note that scale economies and network effects reduce costs and increase value to the benefit of consumers.)
The competitive facts on the ground, however, lend no support to fears of imminent market dominance in the AI ecosystem. The precautionary instinct is at odds with the American economics-oriented, case-and-fact-specific antitrust-enforcement philosophy, which—as Barnett puts it—justifies antitrust prosecutions only when “the enforcer can identify sufficient evidence to infer that a particular practice is actually or likely causing harm to competition.” This careful approach avoids the costs of prosecuting benign conduct that holds the prospect, especially in digital markets and “AI space,” of showering enormous welfare benefits on society.
The threat of imminent litigation that interferes in business planning also may be expected to diminish the incentives for AI innovation, to the great detriment of the U.S. and global economies. The many alliances that are driving AI innovation forward could wither. Barnett’s analysis finds that “preemptive intervention on antitrust grounds [in AI markets] appears to lack a reasonable justification and raises the risk of distorting and harming the market’s future trajectory.”
A recent analysis cited by CCIA estimates that “AI-related products and improvements will contribute $15.7 trillion to the global economy by 2030, including $3.7 trillion to the U.S. economy (14.5% of total estimated GDP).” But those benefits could be reduced significantly to the extent that antitrust prosecutions deter welfare-enhancing investments and business arrangements affecting the AI ecosystem.
A Reasonable Path Forward
President Donald Trump’s July 3 executive orders accompanying the “America’s AI Action Plan” explicitly sought to promote U.S. AI leadership by removing regulatory impediments. This deregulatory approach is a central component of the administration’s AI strategy, which also includes the earlier January 2025 executive order on “Removing Barriers to American Leadership in Artificial Intelligence.”
An overly interventionist antitrust approach to AI alliances would run counter to this important priority. Antitrust enforcers in the U.S. Justice Department (DOJ) and Federal Trade Commission (FTC) may wish to keep this in mind when developing enforcement strategies that affect the AI ecosystem.
This would not be a “do nothing” strategy. It would involve monitoring the ecosystem to spot any new practices that could seriously inhibit or otherwise distort competition (such as creating inefficient barriers to the entry of potential rivals, for example). But it would require compelling evidence of likely or actual competitive harm before acting, rather than relying on “fishing expeditions” driven by purely theoretical concerns about the future. Practices that are not well-understood by enforcers should not be the target of preemptive antitrust action.
A substantial reduction in regulatory barriers, coupled with a cautious facts-based antitrust philosophy dedicated to doing no harm, make sense. This could be the best formula to promote a dynamic world-class U.S. AI sector that enhances economic growth.
