Countries Consider A.I.’s Dangers and Benefits at U.N.

The United Nations on Thursday announced a plan to establish itself as the leading global forum to guide the path and pace of artificial intelligence, a major foray into the raging debate over the future of the rapidly changing technology.

As part of its General Assembly this week, the organization said it was implementing a “global dialogue on artificial intelligence governance,” to assemble ideas and best practices on A.I. governance. The U.N. also said it would form a 40-member panel of scientific experts to synthesize and analyze the research on A.I. risks and opportunities, in the vein of previous similar efforts by the body on climate change and nuclear policy.

To begin the initiative, dozens of U.N. member nations — and a few tech companies, academics and nonprofits — spent a portion of Thursday summarizing their hopes and concerns about A.I.

In short snippets, the speakers extolled the promise of the technology to cure disease, expand food production and accelerate learning. But they also identified risks including mass surveillance, the spread of misinformation, the consumption of energy resources and worsening income gaps among people and nations.

“The future will not be shaped by algorithms alone,” Annalena Baerbock, president of the U.N. General Assembly. “It will be shaped by the choices we make together.”

The U.N. program is an effort to ensure that control of the increasingly powerful and pervasive technology is not left in the hands of a few tech companies and a few countries, led by the United States and China.

Since the explosion of generative A.I. technology with the public launch of ChatGPT in 2022, governments around the globe have wrestled with how and when to use the technology. Predictions about A.I., which can craft humanlike text and realistic images, range from supercharging economic growth and prosperity to turning rogue and bringing the extinction of humanity.

The United States, under the Trump administration, has resisted global efforts to regulate A.I., fearing they might hamper American tech companies in their race to stay ahead of China.

“We totally reject all efforts by international bodies to assert centralized control and global governance of A.I.,” Michael Kratsios, director of White House Office of Science and Technology Policy, said on Wednesday.

As the United States has pulled back, China has increased its support for global A.I. initiatives and cast itself as a champion of developing nations. Speaking on Thursday, Ma Zhaoxu, China’s executive vice minister of foreign affairs, warned that A.I. must not become “a game of the club of wealthy nations” and “tool of hegemony.”

The U.N. has been studying A.I. for the last few years, watching as both investment and anxiety mount. It is now campaigning to be the preferred convener for exchanging ideas, best practices and solutions.

The U.N. dialogue could also result in the formation of an independent A.I. watchdog somewhat like the International Atomic Energy Agency, which under the auspices of the U.N. promotes the peaceful use of nuclear technology and seeks to prevent the spread of nuclear weapons.

Concerns about the trajectory of the technology abound. This week, a group that includes technologists, Nobel laureates and former government leaders called for global A.I. guardrails at the General Assembly session.

More than 200 signatories warned of “unprecedented dangers” ahead as A.I. could soon surpass human intelligence and present socially destabilizing risks, including engineered pandemics, widespread disinformation and mass unemployment.

The signatories appealed to governments to reach agreement by the end of next year on what A.I. should not do, then require that companies put safeguards into place and test those restraints before releasing new products to the public. The group did not propose specific steps, but such a list might well include prohibitions on lethal autonomous weapons and self-replicating A.I. software, which could result in the technology escaping human control.

“We urge your governments to establish clear international boundaries to prevent universally unacceptable risks for A.I.” Maria Ressa, a winner of the Nobel Peace Prize in 2021, told the assembled delegates on Monday,

Technologists and businesses could design the technology to empower workers rather than replace them, said Mr. Acemoglu, a professor at the Massachusetts Institute of Technology. He pointed to earlier digital inventions like spreadsheet software, which democratized financial analysis and increased employment, and internet search, which enhanced the ability of people to find useful information quickly.

But that is not the trajectory Mr. Acemoglu sees at the moment. “The A.I. quest is currently focused on automating a lot of things, sidelining and displacing workers,” he said.