(Yicai) Jan. 9 -- Experts from the Beijing Municipal Public Security Bureau have jointly authored a report suggesting China should establish a comprehensive governance framework to address crimes committed using artificial intelligence technology in light of their rapid increase.

Criminals may commit fraud using AI for deepfakes or voice alteration, which is more authentic than traditional methods and makes it difficult for victims to recognize the schemes, according to the report published in the journal of the Chinese Academy of Sciences on Jan. 7.

In addition, independent criminal acts caused by autonomous driving systems, embodied intelligence in robots, or other AI systems may also occur, the experts noted. In the real world's more complex scenarios, intelligent decision-making agents may violate some basic principles pre-set by humans, differentiating AI crimes from traditional ones.

"In the foreseeable future, the incidence of crimes using AI tech will continue to rise, posing significant challenges for social governance," the experts stressed.

AI deepfake frauds surged 30 times last year from the prior one, with the number of AI-generated phishing emails soaring 11-fold, according to a study by 360 Security Technology, also known as Qihoo 360, cited in the report.

The experts proposed adapting laws and regulations to the development and application of AI technologies, implementing a dynamic legal updating mechanism to periodically revise relevant provisions in response to emerging tech. Accountability mechanisms for AI developers, providers, and users should also be set up while creating systems for detecting and handling AI crimes, conducting regular monitoring and early warning, developing a dynamic classified regulatory mechanism for AI, and enhancing talent training and international cooperation.

Universities should start specialized AI programs focusing on public safety while emphasizing practical case analysis and training and creating a dedicated database for AI crimes for real-time sharing of crime data to effectively deter such cross-border crimes, the experts said.

Editor: Martin Kadiev