The field of artificial intelligence (AI) presents one of today’s greatest opportunities to advance the potential of human achievement and progress, inspiring interest on the international stage regarding the capabilities of this powerful technology. Even Vladimir Putin has declared in regards to AI, “Whoever becomes the leader in this sphere will become the ruler of the world.” The recent advent of ChatGPT, a sophisticated chatbot and trained language model, revolutionized the world of AI and brought its vast potential to the collective forefront. This rapid progress in AI technology has sparked concerns about the world’s readiness to handle its development and use in safe and ethical ways, such as the possibility of dangerous applications in authoritarian settings, most relevantly China, as well as the uncertain risks posed by future AI technologies.
US-China Competition and Chinese AI Development
The United States and China, as the world’s foremost rivals in AI development, top two countries in AI research, and intense technological competitors and political rivals, are poised to become global giants in this revolutionary technology.
China’s AI development is highly concerning to the United States; some in the United States view the countries’ competition as a contest between democracy and authoritarianism—a sentiment highly reminiscent of the Cold War—in which Chinese dominance could pose a challenge to US prominence on the world stage and even to the liberal international order with its associated rights and priorities. Artificial intelligence can greatly bolster economic and military power and, thereby, political ascendancy; AI is expected to contribute US$17.5 trillion to the global economy by 2030. Therefore, military advances driven by AI would be invaluable in out-matching rivals.
For now, it seems likely that the United States will continue to lead in AI, the release of ChatGPT providing a tangible sense of this advantage. Yet, several Chinese tech companies are already on their way to developing comparable language models. China does face some complications in this endeavor; for example, Chinese is less conducive to training a language model than English, and bans are in place to prevent shipping materials from the United States that could aid in AI development. Aware of this limitation, President Xi Jinping declared a commitment to increased self-reliance to better compete.
Elements of China’s authoritarian state make AI more susceptible to usages viewed by the West as serious violations of international law. For example, China’s limited data protection allows for access to a large amount of data. AI-driven data collection could greatly contribute to China’s nascent social credit system, which was published in November 2022 as a draft law but remains stagnant on a large scale. Many argue that the system runs contrary to customary international law, and the West considers it to severely violate human rights and to be a massive loss of privacy, freedom, and justice, though China denies this. Additionally, the use of AI to further increase monitoring of the internet and public areas would lend the Chinese government significant insight into individual patterns and information, furthering censorship, data collection, and control.
There is also a possibility that China will monopolize AI development in a state-owned enterprise (SOE), which would place it entirely under the control of the Chinese government. SOEs, which comprise at least 23 percent of China’s GDP, play a crucial role in both China’s economy and in the government’s ability to maintain power over the country.
The United States, subject to democratic restraints, has less freedom in AI development and applications than China, but that certainly does not mean that US usage of AI is or will be benign. For example, many question its uses in the military and on the US-Mexican border, as well as raising concerns about privacy and human rights in the face of AI-driven facial recognition technology used by police for surveillance and other potential applications. Additionally, the American Civil Liberties Union (ACLU) has pointed out that civil rights protections from AI do not yet exist for citizens of the United States, and has raised concerns about algorithmic bias in AI.
The dangerous ways in which AI can be deployed, particularly in authoritarian settings, clearly demonstrate the need for large-scale regulation of such technologies.
Future Potential of AI
The trajectory of rapid progress in the field of AI will likely continue. One eventual goal of this progress is the development of artificial general intelligence, or AGI. As opposed to narrow AI, which is created to carry out specific tasks, AGI would be able to complete any human task and even exceed human intelligence. Such a technology holds incredible potential to transform society, but also could be very dangerous; for example, it could become misaligned with human goals and work against them, posing serious risks to the future of humanity.
AGI has not yet been developed, and there is a wide range of expectations for when–and if–it will come to fruition. Some experts conjecture a more conservative 50 percent chance of its development by 2099, others are split on whether it would be available before or after 2060, and other varying estimates have been proposed as well. There is even disagreement about whether it will ever be developed. Although AGI remains hypothetical for now, the specter of risk it raises is a reminder that multilateral cooperation will be necessary to manage such technology.
AI on the International Stage
That said, rising AI capabilities have not gone without consideration on a global scale. Firstly, there is a good deal of international academic interest in AI, displayed in settings like the International Conference on Artificial Intelligence and Law (ICAIL), which has been held every two years since 1987. One group dedicated to research is the United Nations Interregional Crime and Justice Research Institute (UNICRI) Center for AI and Robotics, which seeks to limit the malignant use of AI.
However, international organizations relating to AI are currently extremely limited. In fact, the World Artificial Intelligence Organization (WAIO) is the only international organization that exists dedicated to AI. It focuses on five main components—awareness, ethics and security, policy research, implementation, and social impact—and seeks to link various groups such as the private sector, academia, and the international community to best address AI issues.
The majority of international focus relating to AI has thus far been on its usage in military contexts—for example, autonomous weapons systems—and recent progress has been made toward addressing this dimension on an international scale. The first international summit about military AI was held in February 2023, and on February 16, the US State Department released a “Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy,” which highlights the importance of humans ultimately being in control of operations and the obligation for states’ military AI to adhere to international law.
However, one of the greatest challenges faced by international organizations and laws is the difficulty of enforcement. For example, although 60 countries (notably including both China and the United States) recently signed a statement regarding the responsible usage of military AI, the statement is not legally binding and has faced criticism for that, as well as for being limited in scope. This difficulty of enforcement is partially due to states’ unwillingness to legally restrict their own AI usage for fear of giving rivals an edge. This reluctance, part of a larger pattern in international relations and a major obstacle to cooperation, is representative of prisoner’s-dilemma type situations in which states are compelled to not work together and to pursue their own interests, even when cooperation would be much more beneficial for all parties.
Steps Forward
Considering the human rights concerns raised by China’s AI development, the safety questions likely to arise in future AI technologies like AGI, and the current state of unpreparedness on the international stage for large-scale AI, it is crucial that the international community enact a unified response to address the increasing and potential capabilities of artificial intelligence. This response should be undertaken in earnest as AI increases in sophistication and breadth, and consequently, global impact.
Some possible steps are coordination through the United Nations, establishment and support of international organizations to address AI alignment, ethics, and usage, and creation of international laws and regulations. Although international cooperation does face many pitfalls, these difficulties can be mitigated through increased research, oversight, available information, and communication between states, all elements that would be increased through a more coordinated international response. Only by learning how to responsibly handle artificial intelligence will humanity be able to reap the extraordinary benefits of what it has to offer.