A senior Chinese legislative official has offered a fresh glimpse into Beijing’s evolving approach to legislation on artificial intelligence, signaling new priorities while maintaining continuity with earlier policy thinking.
A senior Chinese legislative official has offered a fresh glimpse into Beijing’s evolving approach to legislation on artificial intelligence, signaling new priorities while maintaining continuity with earlier policy thinking.
Wang Hongyu, a division chief at the Research Office of the Legislative Affairs Commission under the National People’s Congress Standing Committee, recently elaborated on his views in an article published in Studies on Intelligent Society, building on a speech he delivered in 2024.
The piece highlights how the legislature is incrementally shaping a regulatory framework for AI, blending new considerations with established priorities as global competition around the technology intensifies.
Wang said rapid advances in AI are not only reshaping economic competitiveness but also redefining defense and security dynamics, making the technology a central factor in national security and great-power rivalry. His comments echo broader concerns among Chinese officials following recent conflicts that have underscored AI’s growing role in warfare.
While flagging risks tied to AI development, Wang warned that its misuse in chemical, biological, radiological and nuclear fields pose existential risks to humanity.
Meanwhile, Wang pointed to the technology’s latest advances such as humanoid robots, warning that such developments could deepen ethical concerns and amplify labor-market disruptions.
“In the long term, legal liability and social-ethical issues arising from embodied AI products such as humanoid robots, as well as job displacement risks caused by AI substitution and the resulting strain on social-security systems warrant close attention,” Wang wrote.
He also flagged “intrinsic” risks in AI systems, including data poisoning, algorithmic opacity and so-called model hallucinations, as well as deepfake risks. The expanding use of AI, he said, also posed challenges for privacy and personal-data protection, decision-making autonomy and safeguards for minors, underscoring the need for tighter oversight.
Legislation approach, priorities
Against that backdrop, Wang said China should craft an AI legal framework grounded in its national conditions and broader strategic competition, rather than mirror approaches taken by the US or the EU.
Policymakers should balance national security with industrial upgrading, protection of citizens’ rights and social equity, using legislation and regulation to build competitive advantages while sustaining innovation, Wang wrote.
At the same time, he cautioned against overstating early-stage risks. While AI will inevitably introduce uncertainties, an excessive focus on safety could hinder technological progress and weaken competitiveness, he said.
Instead, authorities should prioritize flexible use of existing legal tools, including legislative and judicial interpretations, to address emerging issues such as fair use of intellectual property in large-model training.
Wang also called for stepped-up “scenario-based” lawmaking to tackle bottlenecks in industry development. That includes fast-tracked, narrowly targeted legislation and revisions to existing laws in areas where regulation is most urgent. He cited efforts to amend the Civil Aviation Law to support the development of China’s low-altitude economy as one such example.
Although Chinese academics have produced two proposals for China's pending comprehensive AI law, the top legislature indicated earlier no urgency to rush such a law.
In parallel, he reiterated support for allowing qualified local governments to pilot AI-related legislation, a move aimed at accumulating regulatory experience and paving the way for broader national-level frameworks.
Specific legislation
More broadly, Wang emphasized that any new AI legislation should be closely aligned with China’s existing legal regime to ensure consistency and coordination. This includes integration with data privacy, cybersecurity and data security laws, as well as related administrative rules, to avoid duplication or regulatory gaps.
He also underscored the importance of balancing overarching AI legislation with sector-specific rules.
As a general-purpose technology, AI presents risks that vary widely across applications. In fields such as autonomous driving, smart healthcare and embodied robotics, differences in technical characteristics require tailored approaches to risk classification, compliance obligations and liability allocation. Sector-specific legislation should address these nuances, while a broader AI law focuses on common foundational issues.
Finally, Wang said China may update criminal law in line with judicial practice to curb AI-related misconduct, including threats to individual rights, social order and national security.
Source: MLex
