This week, a Chinese artificial intelligence (AI) company, DeepSeek, shocked the tech world with R1, a chatbot reportedly made at a fraction of the cost of its American competitors. President Trump called this debut a “wake-up call”. This advancement shows how formidable a competitor China is to the U.S. AI industry. DeepSeek’s ability to train its underlying large language model (LLM), v3, at such a low cost—approximately $6 million compared to hundreds of millions for American LLMs—demonstrates the need for continued U.S. investment in AI. 

DeepSeek’s success disproves assertions that scale, money, and compute are prerequisites to AI success. As Axios aptly summarized, DeepSeek has eroded the belief that “size is everything” when it comes to tech. This perspective underscores the competitive dynamics in the space, and the urgency for U.S. companies to continue to foster innovation.

In recent weeks, the federal government has taken important steps to support American AI development. China is catching up with the leading U.S. AI models—DeepSeek has already surpassed OpenAI’s ChatGPT on Apple’s App Store. If the U.S. is to maintain its position as the world’s preeminent technological power, the federal government must continue to support investment and a strong business environment for AI innovation.

Investment promotes innovation and creates efficiencies that allow U.S. companies to lower costs and improve products. As we’ve previously written, the proliferation of American AI companies has created a need for unprecedented levels of investment throughout the AI stack. From chips to cloud computing to LLMs, established companies and startups alike are competing to make cheaper, more efficient products. 

As a result, an AI model is 240 times less expensive than it was two years ago. AI companies and investors plan to put an estimated $1 trillion into AI over the next few years in an effort to continue building efficiencies—and therefore lower costs—according to a recent Goldman Sachs report.

Unfortunately, the previous administration’s approach has hindered American companies’ ability to innovate in AI. The Biden administration DOJ’s proposed remedies in its antitrust case against Google, in particular, threaten to derail American AI leadership:

— Ban on AI investments: The proposal would block Google (which has a track record of investing in companies and helping them grow, like Android) from investing in or partnering with AI startups, stifling innovation by making it harder for new companies to secure capital and scale with support from established firms.

— Forced divestiture: Google would be required to divest its AI investments, disrupting startups that rely on its resources, funding, and expertise during a critical time of global AI competition.

— Limits on product integration: This would mean prohibiting Google from integrating its AI products such as AI Overviews in Search and halt progress in AI developments and applications.

— Mandated data sharing: Google would be forced to share user data—potentially with foreign competitors like DeepSeek—jeopardizing U.S. competitiveness and enabling exploitation of sensitive information.

— Access to proprietary features: Competitors could demand access to Google’s search results and AI-generated features, allowing foreign companies to freeload on U.S. innovations.

The U.S. government must continue to support investment and avoid overregulation of the AI industry. The federal government has taken much appreciated steps recently to support American AI development and recognize the need to compete with companies in China. If the American AI industry is to maintain its position as the global leader, the government must take a pro-investment, pro-competition approach.

— At a Stanford symposium last year, AI pioneer Andrew Ng argued that the U.S. government should be “firmly pro-competition and pro-tech [on AI], including Big Tech and small tech and startups.”

— At the same symposium, former Secretary of State Condoleezza Rice cautioned that the government must avoid regulatory stances that could affect U.S. ability to compete with China: “I worry that overregulation… will actually stall our innovation and not allow us to outrace [China].”

With support from investors and the government, American AI innovation can secure U.S. technological preeminence.

— U.S. government funding for foundational AI and computational biology research—through agencies like the NIH, NSF, and DOE—has laid the groundwork for private sector breakthroughs. For example, Google DeepMind’s AlphaFold, which earned a Nobel Prize in Chemistry, builds on U.S.-backed research into protein modeling and AI algorithms.

— Autodesk’s AI-driven generative design technology, applied in partnership with NASA, enabled the development of an interplanetary lander concept. This collaboration resulted in a 35% reduction in the lander’s structural mass, allowing for additional scientific instruments or fuel, which significantly enhances mission efficiency and capability and helps secure U.S. leadership in aerospace technology and space exploration.