Guardrails needed for AI growth

DeepSeek's open-source artificial intelligence system has become popular worldwide since its release in January. (Photo/China Daily)

While embracing the rapid rise of artificial intelligence, deputies to the 14th National People's Congress and members of the 14th National Committee of the Chinese People's Political Consultative Conference last week warned of the need for greater regulation to ensure that the technology is not abused.

With AI such as ChatGPT and China's own DeepSeek rapidly spreading across the globe, its application and regulation were a hot topic at China's largest annual political gathering, with many national lawmakers and political advisers expressing their keen interest in the integration of AI into various industries.

Zhang Yi, a CPPCC National Committee member from Shanghai, said the integration of AI with industries will greatly boost the development of new quality productive forces, and may change the underlying business logic in various traditional areas.

In the film industry, AI-generated images and videos can be used to partially replace traditional filming, and in the retail industry, AI could be linked to back-end systems to achieve unmanned operations, he said.

"We embrace AI. It can enhance the market competitiveness of small and medium-sized enterprises, provide a platform for innovators to practice and create job opportunities," said Zhang, who is also a partner at law firm King &Wood Mallesons.

"However, we've seen some problems caused by the rapid development of AI, such as discriminatory algorithmic decision-making, privacy violations and weak data security," he said, emphasizing that this uncertainty around the technology and associated risks will be a constant challenge for governance.

"It is essential that applications of AI are well-regulated through legislation, as we need to find a balance between high-quality development and high-level security, thereby promoting the healthy growth of the technology," he added.

Lu Lin, a deputy to the 14th NPC who is also a doctor from Shandong province, suggested the continued, cautious integration of AI into healthcare services, adding that regulation on big models should be used to guarantee data security, according to a quote from China News Service.

Kimi, an artificial intelligence system by AI startup Moonshot, is displayed at the Apsara Conference in Hangzhou, Zhejiang province, in September. (Photo/China Daily)

Wider application

Since its release in January, DeepSeek's R1 reasoning model has already been integrated into various sectors in China, including education, healthcare and justice.

On March 4, Lou Qinjian, spokesman for the third session of the 14th NPC, said at a news conference that technological breakthroughs made by Chinese companies such as DeepSeek demonstrate that an innovative, open and sharing approach to improving people's livelihoods is the right choice.

Following DeepSeek's decision to make its latest AI model open source, Lou said it has facilitated the wider application of AI technologies worldwide and contributed China's wisdom to global development.

Zhou Hongyi, a CPPCC National Committee member and founder of Chinese cybersecurity company 360 Security Group, said that 2025 would bring about a huge variety of AI applications, saying their use will make industries more efficient and reduce costs.

Doctor Lu agreed, highlighting AI's use in the healthcare sector. "Large models like DeepSeek, through extensive learning from medical literature and case data, can assist medical staff in making diagnostic and treatment decisions. They can quickly access information across various fields, improving the efficiency and accuracy of medical services," he said.

A visitor asks a question to Doubao, an AI model affiliated with Chinese tech giant ByteDance, during the 2024 Volcano Engine Force Conference in Shanghai in December. (Photo/China Daily)

Rising concerns

Zhang, the lawyer, praised AI for being the core driving force for the latest round of technological and industrial revolution, but said that data security and emerging disputes brought by the technology have raised concerns.

"If AI systems are hacked or infected with viruses, they could endanger personal information, public order and even national security," he said, noting that the construction of an AI security framework should be made a priority.

Zhong Zheng, a deputy to the 14th NPC and vice-president of Chinese home appliance maker Midea Group, told Guangzhou Daily that her concerns focus on the risk of those who use AI to create fake videos that mimic the appearance and voices of celebrities or experts.

"This AI-generated content can mislead and defraud consumers, especially the elderly, causing significant financial losses," the paper quoted her as saying.

In October, Lei Jun, founder and CEO of Chinese tech company Xiaomi, was troubled by a series of videos that used AI to imitate his voice. "I don't think using AI in this way is a good thing," said Lei, who is also an NPC deputy, commenting on the issue.

He wrote on social media that the use of AI for face swapping and voice syntheses will trigger violations of portrait rights and personal information, and even lead to fraud.

The improper use or abuse of AI can cause irreparable damage to the reputation and image of individuals or enterprises, and pose risks to social governance, he added.

Dispute handling

Chinese courts have seen a rising number of lawsuits related to AI as its application spreads.

In April, the Beijing Internet Court heard a case in which a voice-over artist claimed that her voice had been used without her consent in audiobooks circulating online. The voice had been processed by AI.

The plaintiff took five companies to court, including a cultural media enterprise that had provided recordings of her voice for unauthorized use, an AI software developer and a voice-dubbing app operator.

The court found that the cultural media company sent the plaintiff's recordings to the software developer without her permission, and the developer used AI to mimic her voice to create AI-generated products.

Because the AI-powered voice mimicked the plaintiff's vocal characteristics, intonation and pronunciation style to a high degree, it was eventually identified as her voice, the court said, ruling that the behaviors of the cultural media company and the AI software developer infringed upon the woman's voice rights under China's Civil Code.

Earlier, the court concluded China's first case on the copyright of an image generated using AI, identifying the AI-generated picture as an artwork under protection because of the originality and intellectual input of its human creator.

In February, a court in Wuhan, Hubei province, also ruled in favor of a netizen whose AI-powered art was used in an advertisement without her permission.

Wu Yiqin, a CPPCC National Committee member, said that strengthening copyright protection has become a crucial issue in the development of AI.

He likened AI to a double-edged sword, underscoring that it involves not only the rights and interests of a broad range of literary, artistic and technological creators, as well as the standardized and healthy development of the copyright industry, but also the high-quality growth of China's digital economy.

Regulation needed

In response to the increasingly challenging landscape of copyright protection, Wu, who is also deputy head of the China Writers Association, emphasized the necessity of building a forward-looking intellectual property legal framework.

It is urgent to accelerate the research and refinement of copyright protection standards in the field of AI and formulate regulatory legal documents, Wu said.

Zhong, from Midea, called for further improvements to laws and regulations to fundamentally curb fake content generated by AI, saying that current laws and rules regarding AI-generated videos are inadequate, especially in terms of copyright ownership and privacy protection, Guangzhou Daily reported.

To address the issue, she suggested the country amend laws such as the Copyright Law to clarify the liabilities for infringement of AI-generated content, the report added.

Considering the long-term coexistence of humans and AI, Zhang, the lawyer, said that it is a must to seek a balance between the regulation and development of the technology, promoting technological and industrial development in a healthy manner with legislation.

He called for adopting a flexible and prudent legislative strategy in the process of achieving this balance. "For example, it will be suitable to interpret some current laws, or ask regulatory authorities to issue guidance to address some prominent or urgent problems caused by AI," he said.

"When it comes to specific scenarios, pilot programs and regulatory sandboxes can be used to explore solutions. Alternatively, local legislation or departmental rules can also be introduced."

While providing legal support for AI innovators, he noted that corresponding obligations should be imposed on the technological service and product providers through legislation.

AI risk prevention and accountability need to be clarified in legislation, he said, suggesting AI applications be classified based on their risk levels and then subject to categorized regulation.

Zhang emphasized that AI-related legislation should be forward-looking and have an international perspective, as AI is a dynamically evolving technology that has become a key area of international competition.

"Legislative work needs to counter foreign containment, and also establish China's own principles, so as to facilitate the international development of China's AI industry and its participation in the formulation of international rules for AI governance in the future," he added.

Lou, the spokesman for the third session of the 14th NPC, said that the Chinese government has attached great importance to the development of AI and its risk prevention, saying that China has proposed the Global AI Governance Initiative to help bridge the technological divide and prevent sci-tech innovation from becoming only for rich countries and the wealthy.

"China stands ready to work with all countries to promote the healthy development of AI, boost world economic growth and increase the well-being of people all over the world," Lou said.

Action planned

The work report of the NPC Standing Committee, which was submitted to the session on Saturday, said Chinese lawmakers will intensify research on legislation in emerging sectors such as AI, the digital economy and big data.

Ma Yide, a deputy to the NPC, said China must act fast as the world is changing fast with the development of AI. "Digital technologies are reshaping modes of production, lifestyle and governance, and are accelerating the reconfiguration of global resource allocation and the global competitive landscape," he said.

"Therefore, it's inevitable to strengthen the legal framework to support the development of the digital economy," said Ma, who is also a professor at the University of the Chinese Academy of Sciences.

He praised multiple pieces of legislation that have been introduced over the past few years, such as the Personal Information Protection Law and the Data Security Law, but pointed out that the legislation is not systematic enough.

"Issues in the digital domain, including data, personal information and AI, are interconnected and cannot be addressed in isolation," he stressed, suggesting the country conduct a thorough study in this regard and formulate a special digital legislative work plan.

Systematic legislation will provide comprehensive solutions for problems such as personal data protection, copyright protection, data ownership, AI and the release of public data, he added.