ChinAI #222: Five Improvements to China's Generative AI Draft Regulations

Greetings from a world where…

Keegan Murray is a revelation

…As always, the searchable archive of all past issues is here. Please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).

Feature Translation: 5 Recommendations for Improving the Generative AI Draft Regulations

Context: This week’s feature translation provides a window into how debates over China’s recent draft generative AI regulations are playing out. On this topic, a WeChat public account (数字经济与社会 ) that covers digital economy and society convened seven Chinese experts, including scholars at Peking University Law School and the China Academy of Information and Communications Technology. They came up with five recommendations to improve the draft regulations. Big shout out to Matt Sheehan for flagging this article (link to original Chinese).

Key Takeaways: Some suggested revisions seemed to narrow the scope of the regulations to only cover public-facing generative AI systems, possibly reducing the regulatory burden on business-facing products.

• To clarify the regulatory scope, which was their second recommendation, the group of scholars suggested this change to Article 5 (image below)

• Note how the proposed changes outline a narrower definition for generative AI providers: only those who provide the public with services are included.

• Recommendation #5 also highlighted this issue (image below). Again, the proposed changes only require “generative AI service providers with public opinion properties or social mobilization capacities” to perform security assessments.

In their recommendations, the Chinese AI policy experts also emphasized that these generative AI regulations build on previous related measures:

• For instance, the above recommendation #5 ties legal enforcement to the “Internet Information Service Algorithm Recommendation Management Regulations.”

• One of the minor changes adds the Data Security Law to a list of requirements for training data.

Lingering questions: What does the Chinese large language model ecosystem look like in the future? Will these regulations nudge providers to target to-business applications, rather than public-facing platforms like ChatGPT?

FULL TRANSLATION: Five Recommendations for Improving the “Measures for the Management of Generative Artificial Intelligence Services (Draft for Comment)”

ChinAI Links (Four to Forward)

Should-read: GovAI’s Annual Report 2022

Congrats to GovAI, my old intellectual nest, for another excellent year of work. This annual report provides a great summary of the team’s research and field-building work.

Should-read: A Survey of AI Researchers on the Importance of Compute

Very interesting survey findings reported by a CSET team of Micah Musser, Rebecca Gelles, Ronnie Kinoshita, Catherine Aiken, and Andrew Lohn:

“Some researchers express skepticism about the government provision of these resources and concerns about an exclusive focus on scaling up compute.

Our results suggest that compute cannot be viewed as an all-purpose lever for promoting AI progress. Provisioning compute domestically and restricting access to it internationally may promote or constrain certain types of AI research, but may have less impact on other AI research areas. Well-intended attempts to democratize AI research by provisioning large-scale compute may even run the risk of exacerbating existing inequalities in compute use. This report’s findings suggest that in some ways, talent is more important than compute for fostering AI research, so policymakers should evaluate how compute-focused interventions can be coupled with policies to foster AI talent in order to effectively promote AI research progress.”

Should-read: How will China’s Generative AI Regulations Shape the Future? A DigiChina Forum

DigiChina brought together an excellent group of experts to discuss these latest regulations. I found the take by Yan Luo and Xuezi Dan of Covington & Burling extremely helpful. They identified some of the same issues raised above by the feature translation:

“With respect to the scope of application, the draft Measures would regulate generative AI services that are “provided to the public” in mainland China. It is unclear from the wording whether “the public” refers to consumers in China, thus excluding generative AI services offered to enterprises from their scope. It is also unclear whether providers of generative AI outside of China that are not specifically targeting the Chinese market will be subject to these rules. Providers of generative AI services also seem to include both the companies providing underlying technologies and companies offering services at the application level.”

Should-read: How Russia Killed its Tech Industry

For MIT Technology Review, Masha Borak details the decline of Russia’s tech industry: “In the months after the invasion began, Russia saw a mass exodus of IT workers. According to government figures, about 100,000 IT specialists left Russia in 2022, or some 10% of the tech workforce—a number that is likely an underestimate.”

Thank you for reading and engaging.

These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is an Assistant Professor of Political Science at George Washington University.

Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).

Any suggestions or feedback? Let me know at chinainewsletter@gmail.com or on Twitter at @jjding99