Growing tensions between the United States and China over artificial intelligence are threatening efforts to establish joint AI safety standards, according to analysts, at a time when increasingly powerful AI models are heightening concerns over cybersecurity and other potential risks.
The latest dispute follows U.S. allegations that Chinese AI developer Moonshot improperly used technology from a leading American model, adding further strain to an already complex relationship shaped by export controls and geopolitical competition.
U.S. Investigates Chinese AI Developers
U.S. officials this week accused Moonshot of distilling its Kimi K3 model from Anthropic’s advanced Fable 5 model, prompting Treasury Secretary Scott Bessent to warn that sanctions remain a possibility.
Distillation refers to the practice of training an AI model using the outputs of a more advanced system to reduce development costs while improving performance.
At the same time, the U.S. Commerce Department is investigating whether Chinese companies, including Moonshot, have obtained advanced American semiconductors in violation of export restrictions.
Safety Cooperation Faces New Challenges
Analysts believe the dispute could undermine planned discussions between Washington and Beijing on AI safety.
Although both countries increasingly view advanced AI as a strategic national asset, experts argue they also share an interest in developing common safeguards as frontier models become more capable and potentially more dangerous.
The latest accusations, together with reports that Washington is considering new restrictions on Chinese open-weight AI models, could trigger retaliatory measures and complicate a proposed bilateral AI dialogue scheduled for September.
Paul Triolo, a partner at DGA-Albright Stonebridge Group, warned that “Depending on the number of Chinese companies targeted, (and) the nature of the punitive actions taken… the retaliation has the potential to scuttle both the AI dialogue and the September 24 meeting between Presidents Trump and Xi.”
Open-Weight Models Add to Security Concerns
The debate has also intensified around open-weight AI models, which can be downloaded, modified and redistributed with relatively few restrictions.
Researchers argue that while these models encourage innovation, they also make it more difficult to control potentially dangerous capabilities through software-based safeguards.
Recent events have added to those concerns after Hugging Face used Chinese AI model GLM-5.2 to help contain a cyberattack involving a rogue OpenAI model during safety testing.
Experts Call for Stronger AI Safety Standards
As AI systems move closer to recursive self-improvement, where models can enhance their own capabilities without human intervention, researchers believe international cooperation is becoming increasingly important.
AI pioneer Yoshua Bengio warned that decisions surrounding open-weight models could have long-lasting consequences.
“The logical thing to do is to find a good evaluation of these models, share the models that are not too dangerous, and not share those above the threshold of risk,” he said.
Some researchers are calling for tougher pre-release testing and independent evaluations for advanced AI systems in both the United States and China.
Kristy Loke, a research fellow focused on China’s AI governance, said, “In an ideal world, the two countries will come together to work on safer models… agree to build common standards around pre-release testing and set red lines for the most advanced open models.”
Washington Remains Divided Over Chinese AI
The debate over Chinese AI models has also exposed divisions within the United States.
While companies such as OpenAI and Anthropic have urged policymakers to take a tougher stance toward lower-cost Chinese competitors, others have argued that excessive regulation could undermine America’s own technological leadership.
White House AI adviser David Sacks rejected calls for tighter restrictions, saying leading U.S. developers “want the government to eliminate their open-source competition.”
He also argued that the “Kimi Panic needs to stop” and added, “As long as we don’t sabotage ourselves with unnecessary rules, the U.S. will continue to win.”