The United States on Thursday will spearhead a United Nations resolution on artificial intelligence, the international organization’s first on the powerful new technology, as Washington vies for a leading role in the global governance of AI amid increasing pressure to counter China’s rising influence in the field and floundering efforts to regulate it at home.
Members of the United Nations General Assembly are set to vote on an AI resolution Thursday.
The resolution, which was sponsored by the U.S. and will be voted on by the UN General Assembly on Thursday, aims to “promote safe, secure, and trustworthy AI systems” and advance a “shared approach” to the rapidly advancing technology.
The U.S. motion is co-sponsored by more than 50 other countries and urges all 193 member states of the UN to endorse “effective safeguards” for the use of AI and ensure its development is “responsible and inclusive,” according to a draft copy seen by Bloomberg.
It aims to close the digital chasm separating richer and poorer countries and ensure all nations are involved with shaping global governance of AI, according to the Associated Press.
In a statement to the Associated Press, U.S. National Security Advisor Jake Sullivan said the resolution “would represent global support for a baseline set of principles for the development and use of AI” and set out “a path to leverage AI systems for good while managing the risks.”
The text also sets out a “shared vision” of principles that should govern all AI systems, according to a joint statement from the dozens of countries co-sponsoring the resolution, including having a “human-centric” focus, being “reliable, explainable, ethical, inclusive, privacy preserving, and responsible,” as well as being aligned with sustainable development and having a respect for human rights.
The U.S. has said it hopes it will pass with the support of all member states and, if adopted, it will “be the first-ever standalone effort at the UN to establish global consensus on AI,” said U.S. Ambassador to the UN Linda Thomas-Greenfield.
The resolution is not binding on member states and its passing will not oblige countries to toe the line on AI governance or enforce a particular set of rules. Its passing would be important through highlighting a broad agreement on principles and future directions between countries, however, and would apply pressure to countries that elect to eschew the framework. However, the text notably skirts some of the more contentious areas of AI governance—the reported exclusion of military applications is particularly egregious—that are already a heated topic in both industry and between nations. The failure to address these, even in a non-binding manner, leaves several crucial matters on the table that are certain to shape the future of AI development and divide nations in the future. Given the prominent role private companies are playing in developing AI, especially big tech firms like Google, Meta, Amazon, Baidu and OpenAI—who arguably play bigger roles than government in shaping the technology—agreement between states still leaves an opening for a comprehensive approach to dealing with industry. The draft resolution, while mentioning the private sector, appears to be relatively timid in this regard. According to Bloomberg, the draft merely “encourages” the private sector to follow relevant local and international laws on AI.
Washington has been keen to assert itself as a global AI leader in recent years and U.S. diplomats have made a concerted effort to step up and assume a leading role in global AI governance. A key driving force of this is Washington’s desire to counter Beijing and its rising prominence on the global stage, as well as to curb China’s influence on the topic as an AI superpower with capabilities that rival its own. Washington’s efforts to lead global governance of AI contrast starkly with its failure to regulate the industry at home. While there is a patchwork of regulation from existing laws and in individual states, comprehensive federal regulation over the important technology has not yet materialized and does not appear likely to do so in the near future.
The document largely focused on the potential benefits AI could bring, especially to meeting the UN’s failing development goals for 2030. The joint statement from the resolution’s co-sponsors says AI could be used to “address the world’s greatest challenges, including those related to poverty elimination, global health, food security, climate, energy, and education.” Richard Gowan, an analyst at the International Crisis Group, told AFP that “the emphasis on development is a deliberate effort by the US to win goodwill among poorer nations,” adding that “it is easier to talk about how AI can help developing countries progress rather than tackle security and safety topics head-on as a first initiative.”
The U.S. reportedly began negotiating with UN member nations on the AI resolution around three months ago, according to the Associated Press, citing a senior U.S. official. The country has reportedly spent hundreds of hours in direct talks with individual nations and 42 hours negotiating, with input accepted from 120 countries.