The false information produced is more confusing. "AI illusion" brings "serious nonsense"... "AI rumors" that are easy to spread and difficult to prevent and control, posing new challenges to government regulators, Internet platforms, technology developers and all sectors of society. How to curb artificial intelligence from lying? We need technological improvement and regulation, as well as systematic governance. Several interviewed experts expressed the need to improve legal norms and departmental supervision, and promote the formation of a multi-party collaborative governance pattern. Clarify the legal bottom line and enhance the deterrence against "AI rumors". Recently, Zheng, a Guangxi netizen who spread rumors about gas price increases through AI, was administratively punished; A group of online accounts spreading "AI rumors" have been ordered by the cyberspace administration to be banned or closed in accordance with the law and regulations; The accounts that teach users how to use AI to "one click undress" and "change face and body", as well as online stores that openly sell related software, have been jointly punished by departments such as cyberspace and public security... Behind the legal punishment is the continuous improvement and perfection of AI related laws and regulations in recent years. The Administrative Provisions on the Deep Synthesis of Internet Information Services, the Interim Measures for the Management of Generated AI Services, and the Measures for the Identification of Artificial Intelligence Generated Synthetic Content were issued, which clarified the legal bottom line for AI users and platform managers and provided a legal basis for law enforcement. Yang Qingwang, Vice Dean of the Law School of Central South University, stated that the problem of low illegal costs for "AI rumors" still exists, and strict punishment should be imposed on illegal behavior, relevant laws and regulations should be improved, and deterrence should be enhanced. Professor Xu Xiaoke from the School of Journalism and Communication at Beijing Normal University suggests that in high-risk areas where false information is generated and disseminated, legal forms such as judicial interpretations should be used to clarify the boundaries of obligations for all parties and develop scientific and reasonable guidelines for determining AI infringement liability. At the same time, the difficulty in obtaining evidence for "AI rumor" cases is troubling the investigating authorities. To solve the difficulty of obtaining evidence in cases involving 'AI rumors', it is necessary to build an evidence collection and identification system that adapts to the characteristics of AI technology. ”Yang Qingwang suggested that by establishing and improving the judicial recognition rules and standards for case evidence, platform algorithms that do not involve trade secrets and personal privacy should be made transparent, in order to solve the problem of evidence and liability determination. Some local public security organs have explored monitoring and identifying "AI rumors". We work closely with research institutions and enterprises to further improve our monitoring, early warning, and rapid identification capabilities for malicious' deepfake 'information. We collaborate with relevant departments to establish a closed-loop mechanism for online rumor inspection and monitoring, feedback verification, consultation and judgment, clue transfer, and crackdown and disposal. For false information synthesized through AI and other methods, we will handle it in accordance with laws and regulations in a timely manner. ”Yang Yongzhi, a police officer from the Network Security Support Brigade of Shenyang Public Security Bureau in Liaoning Province, stated. Strengthening the technological dam and cutting off the transmission chain of "AI rumors". Some netizens have found that on some online platforms, the prompt of "suspected AI technology generation" does not appear in a timely manner when AI generated images and videos are published, giving time for the spread of false information. In order to maintain their own development and interests, platforms should consciously govern 'AI rumors' and cut off the transmission chain of' AI rumors'. ”Yang Qingwang believes that after discovering AI rumors, platforms should immediately take measures such as delisting, deletion, and debunking. Especially in the face of major events and sensitive topics, they should strengthen the authenticity verification of images and videos and promptly eliminate the negative impact brought by rumors. How to strengthen authenticity verification? Pay attention to feedback from netizens. Establish convenient reporting and complaint channels, and promptly verify and handle suspected rumors reported by netizens. For users who repeatedly use AI to spread rumors, prohibitive measures such as blacklisting can be taken to regulate user behavior. ”Yang Qingwang said. Strong technical support is still needed for AI recognition. ”Li Huaisheng, director of the Institute of Network Law at China University of Political Science and Law, stated that technically, it is possible to identify and block the spread of rumors through methods such as algorithm recognition, big data analysis, and blockchain traceability. For example, using AI systems to detect suspicious content and reduce recommendations through natural language processing and image recognition technology. Regulatory authorities can also introduce a 'dynamic traceability responsibility chain' mechanism, using explicit and implicit content identification, to explore the content backtracking and dynamic accountability of 'AI rumors'. ”Xu Xiaoke said. Improve discernment ability, see the true face of "AI rumors", count your fingers, AI may generate "six fingers"; Searching for the direction of the shadow, AI may hang the sun upside down; Looking at facial expressions, AI generated images may have eye mouth coordination issues... It's difficult to distinguish the authenticity of AI rumors, but clues can still be found. Raising vigilance is the first step in enhancing the ability to distinguish "AI rumors". When you see the words' netizen leaks' or 'exclusive insider information', or discover loopholes or contradictions in the information logic, you should be careful - it is likely to have been tampered with or fabricated. Be cautious in discerning and try to find authoritative information for verification; you can also verify the authenticity of information through cross validation, "said Xu Xiaoke. We also need to enhance legal awareness and prevent the spread of rumors unconsciously. Some people may not have a subjective awareness of spreading rumors and simply post AI synthesized content on the internet, resulting in a large number of shares and negative impacts. ”Li Huaisheng said. Regarding this, Li Huaisheng reminds that AI generated content may have factual deviations, logical loopholes, or outdated issues. During use, a cautious attitude should be maintained, and unverified AI generated content should not be directly publicly released to prevent unintentional misleading information dissemination. In terms of media responsibility, experts believe that media platforms should improve their content review capabilities and verify the authenticity of information through official channels and authoritative sources; Self media practitioners should enhance their legal awareness and publish information truthfully, accurately, and objectively. To prevent and control 'AI rumors', knowledge dissemination is essential. Xu Xiaoke believes that by publishing typical cases to guide people to understand the generation mechanism, dissemination logic, language tricks, and recognition methods of "AI rumors", it can help people see the true face of "AI rumors". The rapid development of AI technology requires learning to use "smart tools" intelligently in order to better serve people's production and life. (New Society)