China Seen Struggling to Adopt AI Rules

China is probably “at least” two or three years behind the U.S. in generative AI, said Samm Sacks, cyber policy fellow at New America and senior fellow at Yale Law School, during a Brookings webinar Wednesday. China appears to be relying on “iterations off of cutting edge research” published in other countries, she said. Chinese officials are struggling with how to balance information control with their desire to lead the world on AI, she said. “AI generates and disseminates information that’s of real concern to the Communist Party leadership,” she said. Cyberspace Administration of China (CCA) has “really been in the driver’s seat” on AI regulation, but the Ministry of Science and Technology is “likely going to have the pen for China’s AI law, which is in the works,” Sacks said. The CCA would likely take a more conservative stance on AI rules than the ministry, she said. Chinese censorship “has a limiting effect on the availability and quality of data,” she said. Another issue for China is U.S. restrictions on access to the most advanced semiconductors, and the Chinese semiconductor industry is “generally several generations behind,” she said. China has been relying on a loophole to use cloud service providers to “rent access” to advanced chips, she said: “We need to watch is this an area where the U.S. government is going to come in and try to close the loophole. How successful will China’s AI ambitions be given these constraints from the U.S.?” Marietje Schaake, Stanford University Cyber Policy Center international policy director, urged flexibility in EU regulations as negotiations continue between the European Council and the European Parliament. “We may not know what will come next, but we do know that something else will come next and generative AI is certainly not the last disruptive iteration” of AI, she said.