Washington and Beijing are weighing the launch of official discussions about artificial intelligence, said people familiar with the matter, as their AI competition threatens to become the arms race of the digital era.

The deliberation comes as the White House and the Chinese government are considering putting AI on the agenda for a summit next week in Beijing between President Trump and Chinese leader Xi Jinping.

Establishing official discussions on AI would mark the start of U.S.-China engagement about the issue under the current Trump administration, reflecting a recognition that the rush to produce more powerful AI models could trigger a crisis neither government has the means to manage. The Biden administration started a dialogue with China, but it yielded limited results, and since then the risks have grown.

What both sides have in mind, the people said, is a recurring set of conversations that could address the risks posed by AI models behaving unexpectedly, autonomous military systems, or attacks by nonstate actors using powerful open-source tools.

Treasury Secretary Scott Bessent is leading the American side on the AI track, said people familiar with the U.S. position. The administration is waiting for Beijing to designate its counterpart, the people said. So far, Chinese vice finance minister Liao Min has been involved in discussions with Washington about setting up such a dialogue, people close to Beijing said.

Ultimately, Trump and Xi will decide whether the AI discussions will be on the formal agenda of their May 14-15 summit.

Liu Pengyu, spokesman for the Chinese Embassy in Washington, said China is ready to engage in communication regarding AI risk mitigation.

Outside the administration, some private-sector analysts are already debating what such discussions might ultimately produce and whether more concrete crisis-management mechanisms could eventually follow, such as an AI hotline for communication between high-level leadership. The U.S. and China already maintain a defense-telephone line and other crisis channels—though Beijing has often been reluctant to use them.

When a U.S. EP-3 reconnaissance aircraft collided with a Chinese fighter in 2001, and again when a Chinese surveillance balloon transited the U.S. in 2023, Chinese officials didn’t answer the line, said Rush Doshi, a scholar at Georgetown University and the Council on Foreign Relations.

“The fundamental question is not whether we create a hotline—which I think has some value—but whether the Chinese will use the hotline, which historically they haven’t done,” said Doshi, who served as the National Security Council’s China director under former President Joe Biden.

The two leaderships previously engaged on AI. At a summit in California in November 2023, Biden and Xi launched a formal U.S.-China AI dialogue. The Biden administration had two priorities at the time, said Doshi, who led the negotiations: to initiate a recurring dialogue, and to reach an understanding that the two governments wouldn’t connect AI to nuclear command and control.

The two sides announced in 2024 that humans, not AI, would retain authority over nuclear-launch decisions.

The dialogue itself fell short of expectations. Beijing put the Ministry of Foreign Affairs in charge rather than a technical body such as the Ministry of Science and Technology, Doshi said, limiting the substance of the exchange.

“If the Chinese were serious about dialogue, I think they would have probably put people who know the technology best in leading roles in their delegation, as we did,” he said, adding that the U.S. side accepted the arrangement: “We didn’t really hold them to that as much as we should have.”

That AI has now climbed to the presidential level for a second time—across two different administrations—reflects a recognition that took root before Trump returned to office that the technology poses shared strategic risks requiring high-level engagement.

In 2023, around the time Henry Kissinger turned 100, Xi invited the former secretary of state to Beijing. Few Americans had wielded more influence in China than Kissinger, whose trip to Beijing in the shadow of the Cold War paved the way for President Nixon’s 1972 visit and the normalization of U.S.-China relations.

Kissinger accepted Xi’s invitation but wanted to be sure AI was on the agenda, said Robert Hormats, who was Kissinger’s senior economic adviser in the 1970s and advised him on the 2023 visit. Kissinger believed AI was an area where China and the U.S. might have common interests or at least be able to reach mutual understandings in key areas, Hormats said.

“The Chinese agreed,” he recalled. Xi and Kissinger discussed the need for a longer-term dialogue on AI-related matters, including regulatory issues and mutual security challenges.

That conversation grew into a continuing nongovernmental channel led on the U.S. side by Craig Mundie, former Microsoft chief research officer, with Chinese counterparts that include participants from Tsinghua University and from major Chinese AI-model companies.

Anja Manuel, executive director of the Aspen Strategy Group who participates in the nongovernmental AI dialogue, said discussions have focused on frontier-model safety and on how to design “guardrail” systems that ensure AI models comply with human laws and intentions even as they grow more capable.

“AI is becoming the operating system of global commerce,” Manuel said. “You really can’t have a trade conversation with the Chinese without talking about AI. It’s going to design drugs, write code, route ships and manage factories.”

China is open to talking with the U.S. about AI safety, said veteran trade expert, Myron Brilliant, who recently met with Chinese officials including Xi’s economic czar, Vice Premier He Lifeng.

“The Chinese side said, ‘Look, yeah, we’re going to compete like heck with the U.S.,” said Brilliant, a senior counselor to DGA Group, an advisory firm. “‘But we also can see merit in enhancing efforts to prevent global shocks, and cyber misuse, so we’re open to dialogue around safety protocols, technical safeguards, and governance if the administration wants it.’”

“Stability—not alignment—is the goal,” Brilliant said.

The frame inside both countries is increasingly familiar from the last Cold War: managing strategic stability through dialogue with a rival you intend to outcompete.

“We did this with the Russians very effectively,” Manuel said. “At least the people who were dealing with nuclear weapons security knew each other and knew how the other side thought, even though they didn’t necessarily see eye to eye.”

Write to Lingling Wei at Lingling.Wei@wsj.com

Washington and Beijing are weighing the launch of official discussions about artificial intelligence, said people familiar with the matter, as their AI competition threatens to become the arms race of the digital era.

The deliberation comes as the White House and the Chinese government are considering putting AI on the agenda for a summit next week in Beijing between President Trump and Chinese leader Xi Jinping.

Establishing official discussions on AI would mark the start of U.S.-China engagement about the issue under the current Trump administration, reflecting a recognition that the rush to produce more powerful AI models could trigger a crisis neither government has the means to manage. The Biden administration started a dialogue with China, but it yielded limited results, and since then the risks have grown.

What both sides have in mind, the people said, is a recurring set of conversations that could address the risks posed by AI models behaving unexpectedly, autonomous military systems, or attacks by nonstate actors using powerful open-source tools.

Treasury Secretary Scott Bessent is leading the American side on the AI track, said people familiar with the U.S. position. The administration is waiting for Beijing to designate its counterpart, the people said. So far, Chinese vice finance minister Liao Min has been involved in discussions with Washington about setting up such a dialogue, people close to Beijing said.

Ultimately, Trump and Xi will decide whether the AI discussions will be on the formal agenda of their May 14-15 summit.

Liu Pengyu, spokesman for the Chinese Embassy in Washington, said China is ready to engage in communication regarding AI risk mitigation.

Outside the administration, some private-sector analysts are already debating what such discussions might ultimately produce and whether more concrete crisis-management mechanisms could eventually follow, such as an AI hotline for communication between high-level leadership. The U.S. and China already maintain a defense-telephone line and other crisis channels—though Beijing has often been reluctant to use them.

When a U.S. EP-3 reconnaissance aircraft collided with a Chinese fighter in 2001, and again when a Chinese surveillance balloon transited the U.S. in 2023, Chinese officials didn’t answer the line, said Rush Doshi, a scholar at Georgetown University and the Council on Foreign Relations.

“The fundamental question is not whether we create a hotline—which I think has some value—but whether the Chinese will use the hotline, which historically they haven’t done,” said Doshi, who served as the National Security Council’s China director under former President Joe Biden.

The two leaderships previously engaged on AI. At a summit in California in November 2023, Biden and Xi launched a formal U.S.-China AI dialogue. The Biden administration had two priorities at the time, said Doshi, who led the negotiations: to initiate a recurring dialogue, and to reach an understanding that the two governments wouldn’t connect AI to nuclear command and control.

The two sides announced in 2024 that humans, not AI, would retain authority over nuclear-launch decisions.

The dialogue itself fell short of expectations. Beijing put the Ministry of Foreign Affairs in charge rather than a technical body such as the Ministry of Science and Technology, Doshi said, limiting the substance of the exchange.

“If the Chinese were serious about dialogue, I think they would have probably put people who know the technology best in leading roles in their delegation, as we did,” he said, adding that the U.S. side accepted the arrangement: “We didn’t really hold them to that as much as we should have.”

That AI has now climbed to the presidential level for a second time—across two different administrations—reflects a recognition that took root before Trump returned to office that the technology poses shared strategic risks requiring high-level engagement.

In 2023, around the time Henry Kissinger turned 100, Xi invited the former secretary of state to Beijing. Few Americans had wielded more influence in China than Kissinger, whose trip to Beijing in the shadow of the Cold War paved the way for President Nixon’s 1972 visit and the normalization of U.S.-China relations.

Kissinger accepted Xi’s invitation but wanted to be sure AI was on the agenda, said Robert Hormats, who was Kissinger’s senior economic adviser in the 1970s and advised him on the 2023 visit. Kissinger believed AI was an area where China and the U.S. might have common interests or at least be able to reach mutual understandings in key areas, Hormats said.

“The Chinese agreed,” he recalled. Xi and Kissinger discussed the need for a longer-term dialogue on AI-related matters, including regulatory issues and mutual security challenges.

That conversation grew into a continuing nongovernmental channel led on the U.S. side by Craig Mundie, former Microsoft chief research officer, with Chinese counterparts that include participants from Tsinghua University and from major Chinese AI-model companies.

Anja Manuel, executive director of the Aspen Strategy Group who participates in the nongovernmental AI dialogue, said discussions have focused on frontier-model safety and on how to design “guardrail” systems that ensure AI models comply with human laws and intentions even as they grow more capable.

“AI is becoming the operating system of global commerce,” Manuel said. “You really can’t have a trade conversation with the Chinese without talking about AI. It’s going to design drugs, write code, route ships and manage factories.”

China is open to talking with the U.S. about AI safety, said veteran trade expert, Myron Brilliant, who recently met with Chinese officials including Xi’s economic czar, Vice Premier He Lifeng.

“The Chinese side said, ‘Look, yeah, we’re going to compete like heck with the U.S.,” said Brilliant, a senior counselor to DGA Group, an advisory firm. “‘But we also can see merit in enhancing efforts to prevent global shocks, and cyber misuse, so we’re open to dialogue around safety protocols, technical safeguards, and governance if the administration wants it.’”

“Stability—not alignment—is the goal,” Brilliant said.

The frame inside both countries is increasingly familiar from the last Cold War: managing strategic stability through dialogue with a rival you intend to outcompete.

“We did this with the Russians very effectively,” Manuel said. “At least the people who were dealing with nuclear weapons security knew each other and knew how the other side thought, even though they didn’t necessarily see eye to eye.”

Write to Lingling Wei at Lingling.Wei@wsj.com