As autonomous military technologies advance rapidly, establishing clear agreements between the United States and China on the operational limits and controls of unmanned combat aircraft has become an urgent priority. Recent developments underscore the challenges posed by increasingly autonomous systems in contested regions such as the South China Sea and Taiwan Strait.
In December 2022, a close encounter between a Chinese J-11 fighter jet and a U.S. RC-135 surveillance plane over the South China Sea drew attention to the risks of manned aerial engagements in sensitive areas. Both sides released footage but disputed responsibility for the incident. However, experts warn such encounters may soon involve uncrewed systems whose decisions are made autonomously without human pilots on board.
The U.S. is expanding its autonomous warfare capabilities. On September 30, U.S. Defense Secretary Pete Hegseth announced the creation of a four-star Autonomous Warfare Command to coordinate the military’s growing use of robotic and unmanned systems. The Pentagon has requested $54.6 billion in funding for fiscal year 2027 to accelerate development and integration of these technologies, a significant increase from $225.9 million previously.
China is advancing along a similar path. Last November, the People’s Liberation Army Air Force unveiled footage of its GJ-11 stealth drone operating alongside a J-20 fighter. State media later reported that a single soldier could control swarms of 200 or more drones, which are capable of autonomously planning flight paths even when jammed.
With greater autonomy comes critical questions about the rules governing these systems: how much authority should be delegated to machines, and how can adherence to operational limits be monitored and verified? The U.S. Air Force has stated that its collaborative combat aircraft—a type of uncrewed fighter jet designed to operate alongside piloted aircraft—will not autonomously deploy weapons. Nonetheless, this stance reflects current policy and could change depending on evolving operational requirements.
In contrast, China has provided limited transparency on its protocols for autonomous weapons systems. This lack of clarity adds complexity to managing risks in potential air confrontations.
During President Xi Jinping’s recent state visit to Washington, both nations agreed to establish a dedicated communication channel for AI-related military incidents and to hold a specialized AI dialogue planned for November, alongside signing a memorandum on military crisis communications. While these measures represent steps toward enhancing transparency, their effectiveness depends on timely and detailed sharing of data regarding autonomous system decisions, especially in emergencies. Previous assessments note that U.S. officials, including former President Donald Trump, have suggested that information sharing with China will have inherent limits.
International efforts to regulate autonomous weapons are also underway. The United Nations Convention on Certain Conventional Weapons is scheduled to hold its seventh review conference from November 16 to 20, where member states will consider whether to begin formal negotiations on rules specifically addressing autonomous systems. A nonbinding declaration agreed upon in September may serve as a foundation for these talks.
Experts indicate that outright bans on autonomous fighters are unnecessary, but mission-specific constraints are vital. Commanders should establish explicit parameters for targets, operating zones, and threat responses. Autonomous systems must be programmed to operate within these constraints even if communications with human operators are lost, either ceasing engagement or requesting further input. Rigorous testing and mutual transparency concerning these safeguards are essential components to build confidence.
However, verifying adherence to these rules poses challenges. Former U.S. Air Force Secretary Frank Kendall has cautioned that safeguards requiring human approval for lethal actions could be bypassed after official demonstrations. He advocates for appointing dedicated officials and organizations responsible for testing and certifying compliance, drawing parallels to standard airworthiness protocols. Adoption of such verification mechanisms by both Washington and Beijing could provide a credible foundation for reporting through their AI-related incidents channel.
As encounters between armed autonomous aircraft increase in frequency over contested waters where human pilots are absent, reliable exchange of operational data—including software logs—will be critical to prevent escalations. The upcoming dialogues between the U.S. and China, along with international forums, offer timely opportunities to set standards and mechanisms essential for managing risks in the era of autonomous warfare.