TL;DR

The China Association for Science and Technology (CAST) has announced a boycott of NeurIPS, one of the world’s most prominent AI research conferences. The move follows NeurIPS banning submissions from US-sanctioned entities, a policy that affects major Chinese technology companies including Huawei and SMIC.

Funding redirected to domestic events

CAST will stop funding members to attend NeurIPS and redirect those resources toward domestic AI conferences. Papers published at NeurIPS will no longer count as qualifying research outputs for CAST-administered funding programmes, though academic impact will still be recognised if evaluated by Chinese academic societies.

NeurIPS said its policy was enacted to comply with US law. The conference has historically been a key venue for Chinese AI researchers — in recent years, papers with Chinese institutional affiliations have accounted for a growing share of acceptances.

The fracturing of AI research

This boycott marks a significant escalation in the separation of Chinese and Western AI research ecosystems. Until now, top-tier conferences like NeurIPS, ICML, and ICLR have served as common ground where researchers from both sides published and collaborated. That shared space is narrowing.

The practical effect extends beyond symbolism. Chinese researchers who previously benchmarked their work against international peers at these events will increasingly publish in parallel venues. Over time, this could create divergent standards for evaluating AI research quality and reduce the cross-pollination that has accelerated progress in the field.

Broader tensions

The boycott comes amid wider friction between the US and China on AI matters. Reuters also reports that China has barred two Manus AI executives from leaving the country while regulators review Meta’s $2 billion (approximately £1.6 billion) acquisition of the company — a reminder that technology transfers between the two nations face scrutiny in both directions.

Looking forward

For UK AI researchers and institutions that collaborate with both Chinese and Western partners, this split creates practical complications. The UK has positioned itself as a bridge between major AI powers — hosting the first global AI Safety Summit at Bletchley Park in 2023. A permanently fractured conference circuit would make that bridging role both harder and more valuable.