AI Competition Is Bigger Than the Military

Talk of an AI “arms race” has dominated discourse on U.S.-China AI competition. From a dedicated Wikipedia page to the words of President Donald Trump’s deputy assistant for technology policy, the analogy is everywhere.

Unfortunately, this framing is wrong.

As I recently argued, calling AI competition an “arms race” is far too winner-takes-all, implying isolated AI development between the United States and China when there are, in fact, vast interconnections and interdependencies between AI development in the two countries. Moreover, an “arms race” framing treats AI as one technology, which is certainly wrong and puts a range of AI applications into a single bucket, each of which has varying properties, sets of training data, and development timelines; they’re also built for very specific tasks. As a result, U.S. policymakers may very well damage AI development, miss out on opportunities for American AI advantage, and mishandle AI risks.

But the problems with calling U.S.-China AI competition an “arms race” go deeper still, and it’s worth zooming in on one of these issues in particular. More specifically, a hyper-focus on the military—or even blurring the lines between the military and everything else, as Rosa Brooks argued several years ago—encourages policymakers, especially those in the national security establishment, to focus almost obsessively on AI’s military applications. In turn, it comes at the expense of the United States’ ability to grapple with how to better develop—and better explore the ethical issues around—non-military AI applications that have the potential to greatly influence economic power and norms around technology and human rights.

Stepping back, it’s clear that we’re in an era of great power competition. As the United States made plain in its 2018 National Defense Strategy, “The central challenge to U.S. prosperity and security is the reemergence of long-term, strategic competition by what the National Security Strategy classifies as revisionist powers,” the strategy reads. “It is increasingly clear that China and Russia want to shape a world consistent with their authoritarian model—gaining veto authority over other nations’ economic, diplomatic, and security decisions.”

Among practices like force projection and the organization of national resources to boost state power, great power competition is also characterized by powerful states, like China and the United States, leveraging other elements of national power, like industrial productivity, science, and technology, in pursuit of their global interests. AI falls squarely in this category. By automating tasks and providing better insight into problems, AI promises to bolster national economies. Reports from the McKinsey Global Institute and Accenture Research and Frontier Economics back this point up, particularly when it comes to developing countries; they predict trillions of dollars in worldwide economic growth by 2030 because of artificial intelligence.

This could have a huge impact on great power competition. In our globally interconnected scientific and economic system, large economies lend a great deal of influence on the global stage. With the United States and China already representing the largest economies in the world, maximizing non-military uses of AI within either respective nation could lead to massive boosts in state power, amplified by China and the United States’ influence on global technology norms.

And yet, even with all the evidence of AI’s effect on economic power, the conversation on an AI “arms race” has centered, literally, on just that—arms. And this is nothing short of a missed opportunity.

For one, it’s in the United States’ own best interest to broaden the American policy focus on artificial intelligence to better include non-military applications. Beijing, while attentive to AI and the military, demonstrates a clear strategic investment in artificial intelligence that extends far beyond war. China’s New Generation Artificial Intelligence Development Plan spans discussion of AI’s role in areas as wide-ranging as industrial transformation, education, pensions, urban operations, judicial services, infrastructure maintenance, and autonomous vehicles. Other documents demonstrate the Chinese government’s ambition to play a key role in international standard-setting around artificial intelligence, too.

At the same time, China has championed digital authoritarianism—broadly, using technology to enhance or enable authoritarian forms of governance—such that it’s imperative for the United States to engage widely on ethical issues across a range of AI applications, setting strong democratic norms around the use of technology in society. As more countries debate the role of technology in our lives, it’s crucial that they don’t fall in line with China on this front.

In addition, AI stands to revolutionize existing institutions like health care, and as disease prediction becomes more accurate or doctor fatigue is minimized through automation, this, too, will affect economic power and aid the United States in technological competition with China. There are important questions to be answered around such issues as data governance—another AI-related area in which Chinese policy appears far more developed than in the U.S.—but it’s almost certain that American AI development will suffer relative to China if the U.S. government remains too focused on AI’s military applications.

To be clear, this isn’t to question the importance of work on AI and the military. There’s a definite need to better understand the implications of artificial intelligence for military capabilities both domestically and in other countries. But there’s also a disproportionate skewing of U.S. policy and attention toward AI’s military applications. Trump may have talked about AI and the economy in the recent Executive Order on American AI development, but that doesn’t necessarily translate into sound policy.

Indeed, government entities like the Congressional Research Service are pumping out papers on artificial intelligence and the U.S. military. The same goes for research institutions, which add to a growing body of work on everything from broad projections on AI and the future of war to specific AI military plans in China and Russia. Meanwhile, op-eds and news articles are published almost weekly, it seems, on similar issues. Still others speculate wildly about the future character of conflict and war in the face of artificially intelligent machines. On the policy side, spending on AI research is heavily concentrated in the Department of Defense.

Even though many AI applications are dual-use in that they have both civilian and military applications, it’s not enough to just have the tech exist; it still takes important organizational changes to effectively adopt and leverage AI in non-military cases. These technologies can greatly strengthen the economy and improve the well-being of lives within the United States and far beyond—and that ought to be a larger point of focus for top U.S. policymakers.