Project Details
Updated 07/13/26 · Provided via application · VerifiedChina currently runs the world's largest mandatory AI disclosure system. The mechanism is that every generative-AI service must 1)submit information about themselves to the regulator, 2)pass a security assessment against binding national standards (right now there are three GB/T standards focused on general safety, data pre training and fine-tuning, and data annotation), and 3) appear in a public registry (算法备案), published in monthly batches of appx. 500 to 800 filings.
Right now very few people outside China read these batches. The English-language coverage (Carnegie, Concordia AI, Stanford DigiChina) and influential newsletters describe the mechanism, but they don’t analyze the filings’ content as data.
Where I stand out is that I read this material in the original. I have already analyzed the January, March and May 2026 batches and the GB/T security standards for a comparative EU-China analysis - it is funded by a BlueDot Rapid Grant (publication will be out Aug to Sep 2026).
For this project, I am suggesting to turn that one-time reading of the filings’ content into a regular ongoing public monitor.
The specific monthly output will include a public English-language brief that (1) characterises the new batch (the provider’s name, which frontier providers appear (the more recent filings include ByteDance/Doubao, Xiaohongshu, Inspur; earlier ones - DeepSeek, Alibaba/Qwen, Zhipu, Baidu, Tencent), what the filings disclose vs what was filed but the regulators did not disclose it); (2) tracks changes in the “binding in practice” technical standards and the draft national AI Law; (3) scores the trajectory against a frontier-safety transparency framework (built from Cotra, Heim and Anderljung during the BlueDot AI Governance course): is China trending toward more or less visibility of dangerous capabilities? Alongside the briefs, I will publish a structured English dataset of frontier-relevant filings (provider, model, role, products, use description, filing number), so that non Chinese speaking researchers in the field can work with the registry
.
Who: just me. I am a lawyer (Head of Legal at a US AI company), I have a law degree from Fudan (taught in Mandarin), and an MSc from LSE in comparative China policy. I read regulatory Chinese fluently. Finally, my method for this proposed project has been already proven to work.
Publication: will publish on LessWrong and / or on my own webpage , and distribute through my EU policy network (includes people who have worked with the EU AI Office, researchers at The Future Society , FLI, Concordia AI - these people are currently currently giving feedback on the comparative grant article i am writing). I will also offer the briefs to to existing China-AI outlets.
Theory of Impact
Updated 07/13/26 · By grantmaking.aiTwo mechanisms.
First, calibration. The EU's GPAI regime is being interpreted right now: enforcement powers start in August 2026, the Code of Practice is new, and the systemic-risk compute threshold is under reconsideration after DeepSeek. How strictly the EU calibrates depends partly on what policymakers believe other frontier jurisdictions require. Today that belief is formed from English summaries of formal text, which miss what China's regime actually demands and discloses in practice. Miscalibration is an x-risk problem in both directions: too little regulation tolerates dangerous opacity, too much pushes frontier development to more permissive jurisdictions and the global safety floor is set by whoever is most permissive. A reliable, primary-source picture of the Chinese regime's operational reality lowers the chance of a high-stakes calibration error during the exact window when the error would be made.
Second, early warning. The frontier-governance literature (Cotra's transparency proposals in particular) argues society needs visibility of capability trajectories to see danger coming. China's registry is one of the few places where frontier-adjacent Chinese AI activity leaves a public, monthly, mandatory trace. Nobody is currently watching it systematically in English. A monitor that flags, for example, new frontier-lab filings, changes in what the standards require to be tested, or catastrophic-risk language moving through binding documents (the GB/T standards already instruct providers to treat with caution AI that may deceive humans, self-replicate or self-modify) gives the safety community a low-cost sensor it does not have.
People
Updated 07/13/26 · By grantmaking.aiTeam Member
Discussion
I think China, the mystery land, still have too much to be shared with the world. I endorse this project to add more objective Chinese data points in the english world.
PS: OCPL at Oxford has a project on AI register in China!
thanks for the endorsement , Zeming, and for pointing at OCPL's guide.
I know the guide pretty well, it's a great static snapshot; my project is the ongoing tracker and the EU-facing gap analysis that a one-time guide can't provide
Private comment. Only shown to approved funders and grant reviewers.
This is a rare case of someone with the actual language skills and legal background to read China's AI filings in the original turning that into an ongoing, usable resource for a field that's mostly been working from English summaries.
I think people's understanding of China is very poor and further understanding of China would be great for improved communication, less scary thoughts of China as the "other", etc.
As such, I am recommending we fully fund the remainder of this ($22000-$2350=$19650).
I just need these two questions answered first.
- Did you receive funding from anywhere since submitting this application, or has the funding ask changed for any other reason?
- Please confirm your commitment to post quarterly updates on how the project is going
Hi Marcus!
I am so happy to read that you find this idea valuable!
-
No haven’t received any funding since I sent the application . The funding ask hasn’t changed.
-
yes happy to post quarterly updates.
This seems like solid foundational work that has the potential to add a lot of downstream value, although the upsides seem like they will be harder to measure initially outside of things like site traffic.