With the explosion of interest in artificial intelligence, policymakers around the world are contending with how best to mitigate against the technology’s novel risks without inhibiting its development. China has consistently declared its ambition to become a world leader in strategic technologies such as AI and quantum computing. It has also announced pioneering policies designed to leverage the power of data to fuel the digital economy.
At the same time, China has moved to implement some of the world’s strictest regulations on AI and data.
These regulations make frequent reference to striking a complementary balance between “security” and “development.” In the fast-moving digital economy however, the reality on the ground is more complex and uncertain, with security often taking precedence. Nonetheless, against the backdrop of a slowing economy, recent events indicate that regulators now appear to be shifting the scales in favor of development in their approach to data and AI governance.
Data Security
In recent years, China has built out a comprehensive data governance framework. This includes a dedicated privacy regulation that carries many similarities with the EU’s GDPR. However, China’s data regulations also contain a distinct focus on security. Restrictions on cross-border data transfer have posed a significant compliance challenge for foreign companies operating in the country, Furthermore, a revised Anti-Espionage Law enacted last summer has broadened the scope of potentially sensitive information, intensifying existing data security concerns.
However, in October 2023, the Cyberspace Administration of China (CAC) released draft rules that exempted certain data export activities from review. This includes internal transfers of employee data as well as data transfers necessary to the performance of certain contracts, such as cross-border shopping and payments. The final version of the rules came into force this March, relieving the compliance burden for a large cross-section of companies exporting lower volumes of personal information.
During Tesla chief Elon Musk’s visit to China this week, an influential industry body declared that Tesla had successfully passed a data security review, opening the door for the electric vehicle maker to roll out its autonomous driving capabilities in the country. This comes after years of scrutiny of Tesla’s data-collection practices in China, including a ban on Tesla vehicles inside government compounds due to concerns over their cameras and sensors. In a meeting between Musk and Chinese Premier Li Qiang, the latter praised Tesla’s operations in China as an “example of success” in China-U.S. cooperation.
After a years-long tightening of data security regulations, these developments indicate a new willingness to achieve a greater balance between national security and the needs of foreign investment.
Artificial Intelligence
After ChatGPT took the world by storm toward the end of 2022, China was one of the first jurisdictions to issue targeted regulations on generative AI. In July 2023, the CAC released the Interim Measures for the Management of Generative AI Services, setting out requirements on both the training processes and outputs of large language models. In line with China’s longstanding approach to strictly managing online content, they require AI services “with the attributes of public opinion or the capacity for social mobilization” to carry out a security assessment and file their algorithms with the CAC before they can be deployed to the public.
As a result, for the first half of 2023, large language models developed by China’s leading internet companies needed to wait for regulatory approval before they could be offered to the public. Furthermore, in August 2023, Apple removed more than a hundred apps offering AI chatbot services from its China store in anticipation of the Interim Measures coming into effect. As U.S. tech companies raced ahead with innovative AI services such as ChatGPT and Sora, this prompted a growing chorus among commentators that strict regulatory controls had inhibited the competitiveness of China’s AI ecosystem.
Similar to data governance, recent developments show a noticeable shift in AI policy discourse. During China’s annual Two Sessions parliamentary meetings in March, the official government work report encouraged the use of AI in industry as part of the prevailing policy priority of unleashing “new quality productive forces” in the manufacturing sector. In the week after the Two Sessions, Premier Li Qiang emphasized the need for greater policy support for AI and to create “a relaxed environment for the development of the AI industry.”
On March 28, the CAC released the full list of 117 generative AI models approved for public use. Furthermore, the final version of a generative AI industry standard was released in February 2024, which did away with some of the strictest security requirements contained in an earlier draft.
An Easier Road Ahead?
These developments attest to a dynamic and responsive approach to digital economy regulation in China. In regard to data governance, a more lenient cross-border transfer regime will boost confidence of foreign investors. For AI, the recent acknowledgement of the need to create a conducive policy environment will likely influence the current drafting of a comprehensive AI law.
Nonetheless, geopolitical challenges and domestic priorities will continue to influence China’s policy decisions. Indeed, China’s loosening of the reins of AI regulations is primarily intended to buttress its domestic capabilities against the backdrop of U.S.-led export restrictions of advanced semiconductors. Despite the recent overtures to Tesla, the CAC recently ordered Apple to remove WhatsApp and Threads from its app store in China due to national security concerns. With TikTok facing a ban or forced sale in the United States, China’s Foreign Ministry warned it would “resolutely safeguard its legitimate rights and interests,” hinting at further punitive action against U.S. companies.
Amid this uncertain geopolitical environment, how China balances competing goals of national security, economic growth, and technological advancement may yet shift.