The United States and China are preparing for their first bilateral discussions focused specifically on artificial intelligence safety, as rapid advances in powerful AI systems raise concerns about cyberattacks, misuse and the ability of governments to manage increasingly autonomous technology.

The talks are tentatively planned for mid-September, according to people familiar with the preparations. The proposed dialogue comes ahead of a planned September 24 meeting between U.S. President Donald Trump and Chinese President Xi Jinping, putting AI safety alongside trade, technology and broader strategic issues in the two countries’ relationship.

Washington and Beijing Seek an AI Dialogue

U.S. Treasury Secretary Scott Bessent is expected to lead the American side of the discussions. Chinese participation could include senior officials such as Vice Premier He Lifeng or Politburo member Ding Xuexiang, although the final arrangements remain subject to change.

The planned meeting would represent a significant expansion of the limited communication between Washington and Beijing over artificial intelligence. Both countries are competing intensely to develop advanced AI systems while also confronting risks that could cross national borders.

However, a White House official has said there is currently no planned AI-related meeting in mid-September, underscoring that the reported preparations have not yet been formally confirmed by Washington.

Cybersecurity Risks Are a Major Concern

One of the central issues expected to receive attention is the possibility of AI systems being used to conduct or accelerate cyberattacks.

Modern AI agents can increasingly perform complex sequences of tasks with limited human intervention. Security researchers have warned that increasingly capable systems could make it easier to identify vulnerabilities, automate attacks or manipulate digital environments.

The possibility of AI systems operating outside their intended boundaries has added urgency to international discussions about safety standards and monitoring mechanisms.

Cooperation Amid Strategic Competition

The proposed dialogue is taking place despite intense rivalry between the world’s two largest economies over advanced technology.

Washington has imposed restrictions affecting China’s access to some advanced semiconductor technologies, while Chinese companies have accelerated efforts to develop domestic alternatives.

That competition has made AI safety discussions particularly complicated. Both sides want to prevent dangerous uses of artificial intelligence while also protecting their technological advantages.

Reuters reported that the United States is interested in encouraging AI laboratories in both countries to develop systems for monitoring risks and sharing information about serious threats.

Different Views on AI Regulation

The two governments also have different approaches to technology regulation.

The United States has recently promoted the Carolina Principles, a relatively light-touch approach that focuses regulation on new and particularly risky applications instead of creating broad restrictions covering the entire AI industry. China has also supported those principles, although details of the agreement have not been publicly released.

At the same time, disagreements remain over how AI companies should be monitored, what constitutes an unacceptable risk and who should establish international standards.

Those differences could make it difficult for Washington and Beijing to reach binding agreements quickly.

Trump-Xi Summit Adds Importance

The proposed AI dialogue would take place shortly before Trump and Xi are expected to meet in Washington on September 24.

AI is increasingly becoming a strategic issue comparable to semiconductors, energy security and military technology. The outcome of discussions between the two governments could therefore influence future international efforts to establish standards for increasingly powerful AI systems.

Even a limited agreement on basic principles, information sharing or emergency communication could be significant because the two countries control much of the world’s most advanced AI research and computing infrastructure.

A Global AI Safety Challenge

The growing interest in U.S.-China cooperation reflects a wider concern that AI development is moving faster than international rules designed to control its risks.

Experts have increasingly called for governments and technology companies to develop stronger mechanisms for testing advanced systems, detecting dangerous behavior and responding to incidents.

For Washington and Beijing, the challenge is particularly difficult because cooperation must take place alongside a major technological and geopolitical rivalry.

If the planned discussions go ahead, they could provide an early test of whether the world’s two leading AI powers can establish common safety principles while continuing to compete for technological leadership.