China’s leader Xi Jinping has his eye on the transformative forces of artificial intelligence to revolutionise the country’s economy and society in the coming decades. But the disruptive, and potentially unforeseen, consequences of this technology may be more than the party-state can stomach.
While there are signs that the leadership is considering loosening the grip in this space, the Chinese Communist Party’s instincts towards overregulation, ideological conformity, and cautious incrementalism could stand in the way of China’s ambitions for global supremacy in AI.
Xi has made AI a strategic priority and wants its development to move quickly. Clearly frustrated by the perceived slow pace of China’s innovation and technological progress, Xi focused the third plenum meeting of the CCP on ways to accelerate his version of ‘Chinese modernisation’. At a politburo study session in January this year, he said he wanted China to ‘break away from the traditional economic growth model and productivity development path’ it has been on for the past 40 years.
To do this, Xi wants China to harness ‘new quality productive forces’ that will drive disruptive breakthroughs and not merely produce incremental improvements on existing technologies. Xi wants to encourage great technological leaps forward that will have a ‘profound impact on global economic and social development and the progress of human civilization’, according to a letter he sent to the 2024 World Intelligence Expo in June this year.
He also sees AI playing a central role in advancing China’s military power by pushing the People’s Liberation Army through multiple stages of military-technological development simultaneously rather than sequentially.
China hopes to acquire ‘leapfrog’ technology, particularly in military AI, which it hopes can change the military balance of power by giving the PLA an overwhelming edge. Debates about the future potential of AI by military analysts in China focus on developing the PLA’s ‘intelligentised warfare’ capabilities, referring to the aim of using AI to control the will of the enemy’s top decision-makers. If successful, AI has the potential to bring about revolutionary changes in the way militaries operate and warfare is conducted—comparable to mechanised warfare, information warfare, and even nuclear warfare.
The party-state apparatus is following suit. While there is, in theory, a ‘whole-of-society’ approach to science and technology development, the various parts of Chinese society in practice have different resources, outlooks, and priorities for how to develop and use AI. Ultimately, it is Xi who is pushing for this technology to be China’s deus ex machina, turbo-charging its economy out of a malaise, transforming its military power and capabilities, and catapulting the Chinese nation to the top of the global order.
Mixed signals are coming from the centre of the political system. In 2017, the State Council called for developing the resources to fuel major breakthroughs in AI by 2025 and to be the global leader in AI by 2030—but gave few details on how to do this. The Council has also urged administrative authorities and courts to adopt a cautious and tolerant regulatory stance toward AI. Following this year’s two session meeting, Premier Li Qiang emphasised the need for greater policy support for AI and to create ‘a relaxed environment for the development of the AI industry’.
Meanwhile, the clock is ticking. Xi sees advances in science and technology as a proxy measurement for China’s overall development. In a speech at the National Science and Technology Awards in June, he noted that, although great progress had been made, ‘innovation capabilities are still relatively weak, some key core technologies are controlled by other countries, and there is a shortage of top scientific and technological talent’.
The twin opportunities and risks of AI create a quandary for the CCP as it moves to put safety measures in place that could act as impediments to progress. Getting the behemoth Chinese party-state to head in one direction has historically been difficult but has been a large part of Xi’s focus over his first two terms in power. Chinese regulations are often described as ‘more reactive than proactive’, and this is generally a theme of Chinese policymaking in all areas. Regulation enforcement can be spotty.
Chinese authorities have begun taking regulatory action, including against deepfakes and harmful recommendation algorithms in 2018. Their top concerns, according to a white paper on AI-generated content published by the China Academy of Information and Communications Technology, are to maintain the CCP’s objectives for online information and network security.
But there are other known risks. PwC estimates that AI and related technologies, such as robots, drones and autonomous vehicles could displace around 26% of existing jobs in China over the next two decades—which would strike at the heart of the implicit social contract the CCP has with the people.
The biggest obstacles to AI success are in fact those the party-state itself has imposed to maintain its control on social stability and to manage disruption and change.
The data regulations and laws China has created, for example, have a heavy focus on national security, meaning protection of the CCP’s position of power must trump the easy flow of huge amounts of data needed for AI development and innovation. Legal restrictions on cross-border data transfers pose a significant compliance challenge for foreign companies operating in the country. Draft regulations on generative AI released in May signalled a shift towards more stringent security standards and regulatory oversight.
Thanks to the CCP’s obsession with information control and ideological conformity, Chinese regulators require AI-generated content to ‘reflect core socialist values and not contain content that subverts state power’, putting the country’s AI aspirations on a collision course with its censorship regime—not to mention precluding future research collaboration and exporting technology to liberal democracies.
Some Chinese companies have been better at threading this needle than others. Researchers from Fudan University created a benchmark to test compliance on different Chinese large language models (LLMs) and determined ByteDance’s ‘Doubao-Pro’ to be the best at responding in ways that were ideologically correct.
The party-state’s obsession with security is a straitjacket it chooses to wear. It is always free to loosen it—and there are some signs that it is at least considering that. But to make Xi’s AI dream a reality, China’s bureaucracy will need to walk a shaky tightrope to maintain the CCP’s ‘two miracles’ of economic development and social stability, while also tolerating the unavoidable disruption that comes with technology-driven transformation.
