NeurIPS Reverses Sanctions Policy After Backlash From Chinese AI Researchers

• NeurIPS initially barred researchers from U.S.-sanctioned Chinese firms, spurring backlash.
• Chinese academic institutions threatened to withdraw support and redirect funding.
• Organizers reversed the decision, citing a ‘miscommunication’ with legal advisors.
• The episode reveals how political tensions are seeping into scientific collaboration.
• Thousands of Chinese researchers contribute annually to NeurIPS, highlighting its global importance.

In a dramatic turn of events, the Neural Information Processing Systems (NeurIPS), the world’s premier artificial intelligence research conference, reversed a controversial policy that would have banned participation from researchers linked to U.S. sanctions lists. The initial policy sparked intense backlash from scientists across the globe, especially in China—a powerhouse in machine learning research. The reversal followed warnings that the policy could irreparably fracture international collaboration in AI development, a field many experts say depends on open knowledge exchange.

What Prompted the NeurIPS Sanctions Controversy?

In mid-March, ahead of the 2026 NeurIPS conference cycle, organizers added new language to the event’s submission handbook restricting access to individuals affiliated with organizations on the U.S. Department of Treasury’s Specially Designated Nationals (SDN) list or the Bureau of Industry and Security’s Entity List. These lists typically target governments, corporations, and organizations involved in activities contrary to U.S. national security interests, such as arms proliferation and terrorism—but also include major Chinese tech companies like Huawei and Tencent, alleged to have ties to the People’s Liberation Army.

However, the inclusion of academic publishing under these restrictions raised eyebrows among scholars, as such constraints don’t traditionally apply to research conferences. Although businesses may face export controls under U.S. law, academic journals and peer-reviewed venues generally operate independently of these regulations.

Backlash Grows Among Global Researchers

The policy’s announcement triggered immediate condemnation from both Chinese and international AI communities. Hundreds of researchers voiced opposition online, with several prominent figures announcing they would no longer serve as area chairs or reviewers for NeurIPS.

“I have served as [area chair] for NeurIPS every year since 2020. Just declined,” said Nan Jiang, a machine learning researcher at the University of Illinois Urbana-Champaign in a widely shared social media post.

Others echoed similar sentiments. Yasin Abbasi-Yadkori, a researcher at the AI firm Sapient Intelligence, noted, “That’s one less area chair responsibility for me. If I hadn’t already committed to colleagues, I wouldn’t submit a paper this year either.”

Why This Matters: AI Diplomacy Under Pressure

The episode underscores a larger trend: the growing convergence of artificial intelligence innovation and geopolitical rivalry. Once seen purely as a domain of academic pursuit and technological advancement, AI is now heavily scrutinized through the lens of national security. With its potential applications in defense systems, surveillance, and autonomous weapons, machine learning has become central to strategic competition between superpowers like the United States and China.

U.S.-China Tech Decoupling and Its Ripple Effects

Washington has increasingly moved toward limiting high-tech exports and collaborations involving China, particularly in semiconductors and quantum computing. However, applying such logic to open-access research platforms risks undermining foundational principles of scientific inquiry.

Paul Triolo, a partner at DGA-Albright Stonebridge and expert on U.S.-China tech dynamics, warned that isolating Chinese talent could weaken America’s own standing in global AI leadership. “At some level now it is going to be hard to keep basic AI research out of the [political] picture,” he stated, emphasizing that mutual engagement benefits both nations technologically and economically.

Role of CAST in Amplifying Academic Boycott

The China Association for Science and Technology (CAST), an influential, state-linked umbrella group representing millions of Chinese scientists, issued a formal response to the NeurIPS sanctions policy. CAST announced it would cease funding travel for Chinese scholars attending NeurIPS and instead allocate resources to domestic conferences that it deemed supportive of local researchers' rights.

Additionally, the organization declared it would stop recognizing NeurIPS publications as qualifying contributions for academic advancement and grant eligibility—an effective blow to the prestige value of the conference within China’s scholarly ecosystem.

A Clarification—and Possible Damage Control

After days of escalating criticism, NeurIPS organizers issued a statement clarifying that their handbook had incorrectly referenced a broad federal sanctions database, rather than the narrower SDN list mandated for compliance purposes.

“In preparing the NeurIPS 2026 handbook, we included a link to a US government sanctions tool that covers a significantly broader set of restrictions than those NeurIPS is actually required to follow,” the NeurIPS Foundation explained in a Friday statement. “This error was due to miscommunication between the NeurIPS Foundation and our legal team.”

Despite this correction, questions remain over whether trust can be rebuilt quickly enough to prevent lasting harm to the conference’s reputation, especially among non-Western contributors who feel increasingly marginalized.

Impact on NeurIPS and International Scientific Exchange

NeurIPS has historically drawn significant participation from Chinese researchers, whose contributions account for nearly half of all papers submitted annually. In 2025 alone, Tsinghua University—the top-ranked academic institution in China—was credited on 390 NeurIPS papers, surpassing even industry giants like Google DeepMind and Microsoft Research.

Moreover, researchers from Alibaba were awarded one of the conference’s highest honors in 2025 for developing Qwen, a powerful open-source large language model. China’s increasing output of top-tier AI research makes its exclusion not just diplomatically sensitive, but scientifically counterproductive.

Open Collaboration Versus Techno-Nationalism

While bilateral tensions continue to rise, evidence suggests that cross-border cooperation remains essential in advancing AI capabilities. According to a previous WIRED investigation, joint U.S.-China authorship on NeurIPS papers continues at relatively robust levels, despite political headwinds. Yet incidents like this risk eroding those connections incrementally, pushing innovation into separate silos.

“NeurIPS’ prosperity comes from the joint efforts of researchers worldwide, and its growth and success have long been supported by sponsorships from some of the sanctioned entities too,” commented Yuliang Xiu, an assistant professor in digital graphics at Westlake University in China, after declining his invitation to serve as an area chair.

What Comes Next for AI Conferences and Global Research?

With trust shaken and uncertainty mounting, NeurIPS faces renewed scrutiny—not only regarding future policy decisions but also concerning how transparently it communicates them. Organizers must balance legal obligations against the need to preserve the inclusive ethos that made NeurIPS the gold standard in AI dissemination.

Should similar policies resurface, rival conferences—either in China or elsewhere—may gain traction among disillusioned researchers. Already, calls are growing louder in China for investment in homegrown forums and expanded technical independence.