# SemiAnalysis reports Chinese AI labs rarely publish safety results

Digest AI · Policy & Regulation · published 2026-10-08T17:46:05Z

Canonical: https://digestai.news/story/semianalysis-reports-chinese-ai-labs-rarely-publish-safety-results

## Summary

SemiAnalysis published a detailed analysis on 8 October 2026, arguing that while China’s government rhetoric on AI safety is increasingly explicit, its actual regulatory and industry practices prioritize speed over frontier risk management. The report highlights a disconnect between Beijing’s stated commitments, such as the new AI Safety Governance Framework 3.0, and the lack of mandatory safety evaluations for high-capability models. Unlike the EU or California, China’s current rules focus on content labeling and application-layer security rather than training compute or dangerous capability testing.

The core of the report is a systematic dataset of 857 model releases from nine major Chinese developers (including ByteDance, Alibaba, DeepSeek, and Zhipu) between 2021 and September 2026. The analysis found that only 3.6% of these releases were accompanied by any published safety evaluation from the developer. Just 1.1% had results available at or before the model’s launch. The remaining 94.9% shipped with no safety disclosure at all. This trend has accelerated as release cadence has increased thirty-fold since 2023, with open-weight models making up more than half of recent releases.

The report also reviewed 65 public statements from lab leaders, finding that safety engagement is sparse and concentrated primarily in Zhipu (Z.ai). Other major players, including hyperscalers like Alibaba and Tencent, have rarely published quantitative safety findings for their frontier models. SemiAnalysis concludes that Chinese labs are in a release and price war, and that current compliance mechanisms do not ensure frontier safety, as a lab can meet all regulatory requirements without ever running dangerous-capability evaluations.

## Key points

- Only 3.6% of 857 Chinese model releases had published safety results from developers.
- 94.9% of releases from nine major Chinese labs shipped with no safety disclosure.
- China’s regulations focus on content and applications, not frontier model capabilities.

## Why it matters

This analysis suggests a significant gap between China’s stated AI safety goals and actual industry practices. If major Chinese labs are not conducting or publishing frontier safety evaluations, it complicates international efforts to manage AI risks and may influence US policy debates on export controls and competition.

## Sources

1. [Beijing Will Not Pace the Frontier: China’s Speed-First AI Safety Regime](https://newsletter.semianalysis.com/p/beijing-will-not-pace-the-frontier) (SemiAnalysis, 2026-10-08)

## Cite

Digest AI, "SemiAnalysis reports Chinese AI labs rarely publish safety results", 8 October 2026, https://digestai.news/story/semianalysis-reports-chinese-ai-labs-rarely-publish-safety-results

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/semianalysis-reports-chinese-ai-labs-rarely-publish-safety-results.json