Concerns about an artificial-intelligence (AI) arms race between advanced militaries have been raised by many experts and analysts, including SpaceX founder Elon Musk and United Nations Secretary-General António Guterres. Scientists and engineers are also becoming increasingly vocal in discussions on the weaponization of AI, urging governments to develop regulations to prohibit certain applications of AI in weapons technology. But they are not modern-day Chicken Littles; when they say the sky is falling, listen to them. Many of their fears are based on a deep understanding of the technological capabilities and limitations of autonomous weapons systems and current trends in warfare.
Published in The Ploughshares Monitor Volume 39 Issue 4 Winter 2018 by Branka Marijan
How strange, then, that the conclusion reached by some is that more money should be put into the weaponization of emerging technologies, thus making such a race more likely.
New technologies appeal to militaries that fear being caught off-guard by their adversaries. One-up-manship abounds—each needs to get the latest advancement before the other. In this way, technologies escalate and the danger increases.
Old wine in new bottles
Among the countries that are making significant investments in AI defence/military technologies are China, the United States, Russia, Israel, India, the United Kingdom, and France. This past November, the UK and Australia held separate “Autonomous Warrior” exercises. Those in the UK were the “biggest military robot exercises in UK history,” with 70 systems tested.
The current competition between the United States and China is garnering particular attention from analysts, reminding some of the Cold War. For example, former U.S. Deputy Defense Secretary Robert Work saw the revelation of China’s national AI strategy as a “Sputnik moment”—referring to the Soviet satellite that sparked the Space Race, a feature of the Cold War.
Then, as now, the belief was that superior weapons make superior militaries. This belief is at the heart of the current push to use AI in weapons systems. And it’s dangerous.
In “Beyond the AI Arms Race: America, China, and the Dangers of Zero-Sum Thinking,”(Foreign Affairs, November 2018), Remco Zwetsloot, Helen Toner, and Jeffrey Ding write that “a full return to a Cold War mindset comes with many significant risks, such as prematurely deploying accident-prone weapons systems or inadvertently proliferating dangerous technological capabilities.”
Not everyone wants to focus on the contests among the major powers. Michael C. Horowitz, a professor at the University of Pennsylvania, argues that what we are seeing is not a single AI arms race, but many races. He believes that a focus on competition between the United States and China skews our view of the diffusion of AI technology. As AI is software that can have civilian and military uses, it cannot be long contained within state borders or kept secret.
Horowitz also stresses the extent to which the private sector is investing in and driving the development of AI technologies. As he points out, this “reverses the dynamic from the Cold War, when government investments led to private sector innovation and produced technologies such as GPS and the internet.”
Still, it is difficult to overestimate the significance of the competition among the United States, China, and Russia as a driver in all narratives about the weaponization of AI. And in examining the rationale of these players, it is hard not to hear the echo of the old rhetoric of the Cold War.
The dragon in the room
China is now commonly seen as the awakening technological dragon. Much has been written about the advantages China has in developing AI capabilities. With its large population and authoritarian governance system, China can access incredible amounts of private data.
This Chinese resurgence should be seen in the context of rising trade tensions with the United States. In U.S. media and policy discussions, China is increasingly portrayed as a dangerous adversary. Indeed, President Donald Trump recently accused China of seeking to meddle in midterm elections. Last month, the U.S. Department of Commerce proposed a rule on export controls of emerging technologies that are linked to national security interests. Many see this as an attempt to slow down Chinese advances.
Not all the bellicose rhetoric and claims of technological breakthroughs should be believed, however. According to Harvard PhD candidate Elsa B. Kania, there is still a lot of cooperation among AI researchers, including those in the United States and China. And there are questions about how advancements being made in China, a closed society, actually rate in relation to those of the more open United States and other Western countries. Indeed, experts on China are emphatic that there is much hype about China and a general lack of awareness of the intertwined relationships among Chinese and Western tech firms.
Changing the conversation
The prospect of pouring immense amounts of money into untested autonomous weapons systems controlled by adversarial military and political leaders invested in zero-sum solutions to conflict is, frankly, terrifying. But an AI arms race is being seen as an inevitable reality by more experts. And to what end? As Peter W. Singer, author of Wired for War: The Robotics Revolution and Conflict in the 21st Century, says, “Arms races are not just about the buildup in search of security, they’re about the irony that you end up feeling less secure because of the competition.”
But an AI-fuelled third world war is not inevitable. Efforts at the United Nations to prohibit the use and development of autonomous weapons must be supported by all countries. We need to break the weaponization cycle and reframe the discussion on the uses of AI and emerging technologies in warfare. Ensuring that countries comply with existing arms-control arrangements and develop new ones to regulate emerging weapons technologies is in everyone’s interest. This requires more diplomacy and less inflamed political rhetoric.
And middle-power countries, like Canada, need to decide how they will shape the discussions on emerging technologies and use their voices. In this new arms race, smaller but technologically advanced nations are important players. Canada’s current strategy of remaining on the sidelines in international discussions means that other countries could ultimately determine how new technologies are used. Fortunately, there is still time to change the current narratives.