Artificial intelligence (AI) is everywhere. The concepts underlying it have been in development for decades and are already supporting commonplace tools like Google Translate, Amazon Alexa, and IBM Watson.
AI can be understood as the ability of machines to ‘think’ and make decisions. Much of this has to do with machine learning and recognizing patterns in massive amounts of data. Increasingly, algorithms are being developed and combined to form complex neural networks. These neural networks can be used for different purposes – from natural language processing that enables chatbots to respond to your online queries to computer vision that enables facial recognition to unlock your mobile phone.
One side of the global discussion about AI has been framed in the context of public interest. For example, industry and academic researchers are using machine learning to reduce homelessness, better inform healthcare decisions, and improve emergency planning for natural disasters. Others have also written about the benefits of AI for the public sector, including much-needed assistance for government agencies that commonly lack the resources and staff needed to analyze high volumes of data collected every day.
Beyond the benevolent pursuits of ‘AI for public good’, however, lies a global security dialogue about an emerging ‘AI arms race’. Artificial intelligence can enable the use of autonomous weapons, such as weaponized drones, and broaden the impact of cyber attacks. AI can also be used to improve the reliability of intelligence operations through more advanced data analysis. For these reasons, artificial intelligence has been described by Auslin as ‘the new frontier of military competition’. But the so-called ‘AI arms race’ should not be compared to the military arms race of the twentieth century. As Horowitz observes: “Because AI is a general purpose technology—more like the combustion engine or electricity than a weapon—the competition to develop it will be broad, and the line between its civilian and military uses will be blurry.”
This race to develop artificial intelligence is already transforming transatlantic relations. The United States (US) is considered an AI ‘superpower’ given its strong lineup of tech giants, startups, and investors. The European Union (EU) has weaker industry commitments but is taking a major leadership role in the global discussion about the ethical and socio-economic implications of AI. In April 2018, 25 European countries signed a declaration of cooperation on artificial intelligence. In September 2018, the European Parliament adopted a provisional resolution on autonomous weapons systems, reaffirming its commitment to addressing the ethical questions of weaponized drones and other ‘weapon systems without meaningful human control’. The European Commission and its member states also published a coordinated action plan on AI development in December 2018. Recent Brexit negotiations have raised questions about the role of the United Kingdom, which has become one of Europe’s most advanced AI ecosystems according to a September 2018 report. The EU appears committed, however, to leading an active dialogue. Much like the General Data Protection Regulation (GDPR), which came into effect in May 2018, the European Union’s approach to addressing the ethical questions of AI is expected to become a de facto global standard in the absence of strong policy leadership from the US or from multilateral organizations like the United Nations.
The US government has made marginal progress in developing official policies and strategies to guide the development of AI. According to President Trump’s deputy assistant for technology policy, “America has been the global leader in AI.” Behind the scenes, the White House Office of Science and Technology Policy (OSTP), Office of American Innovation, and the Emerging Technologies Project at the General Services Administration have been isolated from global policy arenas. The Department of Defense, which has launched several AI initiatives, is moving forward without clear government-wide coordination. Some analysts have emphasized the need for using a ‘whole of government’ frame for AI in the United States. Others are calling for a comprehensive, American national artificial intelligence strategy to proactively spur the development of AI and strengthen national security.
Financial investments supporting the advancement of AI are also notable. The European Commission recently approved an investment of $1.8 billion in artificial intelligence by 2020. Other investments are expected to come from individual European states and the private sector. For example, the German government announced plans to make over $3 billion available to the private sector to support AI development through 2025. Similarly, the French government pledged $1.5 billion for artificial intelligence programs. The US government has had lower levels of investment. To date, private sector companies in the US have invested nearly ten times as much in artificial intelligence as the government has in North America. Horowitz concedes, “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.”
China and Russia are also critical players to consider. Recent reports estimate that the Chinese government has already committed over $2 billion to AI development. Like the United States, private sector companies in China like Alibaba are being tapped to bolster Chinese national efforts. The Russian government has also expressed its interests in artificial intelligence but has not yet made comparable investments. The transatlantic community would be smart to work together to prepare for these looming AI power shifts.
Unlike the nuclear arms buildup and the space race of the twentieth century, however, the development of artificial intelligence is being driven by industry and can simultaneously harm and advance the public interest. While US companies may be at the forefront of technology development, it is likely that Europe will lead the global policy dialogue and America will lag behind. What remains to be seen is how this growing AI asymmetry will play out in the global arena.