This regulation applies to individual researchers and institutions, including universities, medical and health centres.
The guidelines set out a general framework for the use of AI, but do not provide detailed rules for specific situations.
One of the main concerns is the use of AI-generated content, which cannot be listed as co-author under the new rules.
Some scientists have cited AI tools such as ChatGPT as co-authors, but this has already been discontinued by many journals.
However, this has led to debates about whether AI should be given credit for discovering new materials or drugs, and whether AI should be considered equal to humans in the pursuit of knowledge. It has even sparked a broader philosophical debate about.
Four people arrested in China for ransomware using ChatGPT
Four people arrested in China for ransomware using ChatGPT
Under the guidelines, generative AI can still be used for research, but content and findings that use the technology must be clearly labeled as such.
Wen Xiaoqing, an associate professor at Shanghai’s Fudan University, said the regulatory impact would be “limited,” at least in his field of scientific archaeology, because generative AI “tends to produce low-quality project application proposals.” He said that there is a high possibility that this is the case. Few people will use it.
A biologist at the Chinese Academy of Sciences also said on condition of anonymity that they would never use AI to write proposals because these language models produce “beautiful but essentially meaningless words.”
However, another researcher at the Chinese Academy of Social Sciences said the spread of generative AI is an “irreversible trend” and that banning it completely would be unrealistic and unreasonable.
The scientist also spoke on condition of anonymity. Researchers will continue to use it for research and grant applications if they think it’s useful to them, but “we need to think about how we can get researchers to use it in an appropriate way.” That’s more important,” he said. And…to keep up with international trends. ”
A Ministry of Science document explaining the new rules says the rapid development of AI technology has “prompted a major change in the paradigm of scientific research.”
However, he continued, the technology raises issues regarding authorship, intellectual property, and data processing, so guidelines are needed.
The ministry said the guidelines were “based on positive experience abroad and reflect international practice.”
It said the guidelines are based on “broad consensus” among the scientific community and will be updated and adjusted based on future technological developments.
Regulation of the use of AI became a major focus of the Chinese government last year, both domestically and internationally.
Chinese government launches AI platform to meet growing demand for domestic computing power
Chinese government launches AI platform to meet growing demand for domestic computing power
At a Politburo meeting in April, he said, while calling the development of AI “very important,” he also emphasized the importance of risk prevention.
According to the Cyberspace Administration of China, these measures do not apply to scientific research institutions unless they provide generative AI services to Chinese citizens.
China’s new scientific guidelines expand on the interim measures. Apart from banning the direct use of AI when creating applications, the rules also require AI-generated content and results to be labeled in text “especially if they contain important content such as facts or opinions.” It also states that there is.
AI-generated content should also be identified in the footnotes, methods section, or appendix of the research paper, along with a description of how it was created and the software used.
The guidelines state that content marked as generated by AI should not be treated as original work and that “an explanation should be provided” if other authors wish to cite this content.
It also says that references generated by AI cannot be used unless they are first verified. University College London’s library guide previously warned that generative AI could fabricate references and citations.
A scientist at the Chinese Academy of Sciences who studies organic chemistry said that restrictions on AI-generated content are necessary because many graduate students are now using AI to write their papers. “It will lead to destruction,” he said.
The scientist, who requested anonymity, believes such tools can make students lazy and make it difficult to develop independent thinking skills.