How DOJ’s Proposal to Break Up Google Would Hurt U.S. Competitiveness in AI

Last October, the U.S. Department of Justice (DOJ) proposed a sweeping set of remedies in response to an earlier court ruling that Google violated antitrust laws with its search business. While most attention has focused on the potential partial breakup of Google—the DOJ has proposed the divesture of the Chrome web browser and the Android mobile operating system—the proposed remedies would also have significant implications for U.S. competitiveness in AI. Given the importance of U.S. leadership in AI to both its economy and national security, regulators should start paying more attention to the potential impact of their proposals on AI.

The DOJ’s proposal would require Google to divest from any investment in a company that makes AI products. In addition, it would prohibit Google from partnering, collaborating, or acquiring any interest in a company that makes “query-based” AI products. The impact of these requirements would be harmful for Google, American consumers, and U.S. competitiveness in AI.

First, it would hamstring Google at an important technological inflection point since AI is booming. The DOJ defines “AI products” as “any application, service, feature, tool, or functionality that involves artificial intelligence capabilities.” This definition is extraordinarily broad, and given the growing prevalence and prominence of AI in virtually every segment of the economy, few companies that Google might invest in are going to be completely untouched by AI. In essence, Google would be unable to incorporate outside AI innovations into its products and services through a typical acquisition or strategic partnership. For example, it could not acquire an AI video generator service and integrate it into Google Slides or an AI text editing service and integrate it into Gmail and Google Docs. At a time when many tech companies are competing based on how effectively they can integrate AI into their products and services, this restriction would unreasonably sideline one of America’s top tech companies.

Second, these restrictions would hurt American consumers. Acquisitions and partnerships between large tech companies and smaller ones are useful because they enable smaller companies to bring their products to market quickly. Marketing a new service to millions of potential users is a much longer and harder path than simply having it integrated into a tool users are already familiar with. Indeed, many startups develop a valuable product or service, but do not have the capabilities or vision to turn it into a successful, scalable business. Denying a company like Google the ability to acquire AI companies would likely keep valuable, productivity-enhancing technology out of the hands of millions of Americans. For example, the restrictions could prevent Google from acquiring an AI-enabled cybersecurity product that could improve information security for millions of its users.

Third, limiting Google from making AI investments would hurt U.S. competitiveness in AI. Investments and partnership with large tech companies have allowed U.S. tech startups to thrive, particularly in AI where many companies face significant operating costs. For example, Google has invested $2 billion in Anthropic in 2023 and $2.7 billion for Character.AI in 2024. Notably, both deals involved non-exclusive agreements: Google does not have exclusive access to either company’s AI models, and the startups are free to use other cloud providers. Restricting Google from investing in these types of AI startups would be detrimental to the U.S. AI startup ecosystem by removing a significant source of funding.

Finally, a separate provision in the DOJ’s proposal would restrict Google from entering into any exclusive agreements with publishers to license their data. This prohibition could prevent Google from funding data gathering initiatives that could have positive social value. For example, Google could not provide museums funding to digitize historic records or cultural artifacts on the condition that the company receive temporary exclusive access to some of the data—a condition that might allow it to see a return on its investment and thus fund more projects. These types of projects can be beneficial for all parties. For example, the Vatican recently partnered with Microsoft to create a digital twin of St. Peter’s Basilica, collecting 22 terabytes of data from cameras, laser scanners, and drones. Prohibiting Google from making these types of investments would not create more competition, it would create a poorer data ecosystem, a key ingredient for AI.

The DOJ’s proposal is a terrible idea for many reasons, but its impact on American competitiveness in AI would be a unique setback at a time when U.S. policymakers should be looking for every potential advantage for U.S. businesses to succeed globally. Fortunately, President Trump has already expressed skepticism about a potential Google breakup, and one of the priorities for his new team at DOJ should be to significantly alter the course of this case in the months ahead.