In recent years, artificial intelligence (AI) technology has developed rapidly and become a hot topic in the global technology field. AI has evolved from pure academic research to an important force in changing human lifestyles. However, some people with ulterior motives have turned it into a tool for spreading rumors and accumulating wealth. In June this year, a short video of "a 70 year old man in Huayang Town, Wuhua County, Guangdong Province was beaten to coma, his grandson jumped into the river to commit suicide, and the batter was sentenced to nine years and ten months in the first trial" spread on the Internet platform, causing a large number of netizens' attention and discussion. After investigation by the public security network security department, it was found that the graphic and textual content of the video was fabricated online rumors that were fabricated arbitrarily. According to the rumor monger Ren, this video, which is several minutes long and full of sound and sound, was all synthesized by AI. At present, the public security organs have lawfully pursued legal responsibility against Ren. This incident is not an isolated case. With the increasing application of AI technology, some netizens use AI technology to create "fake news" and "fake news" in order to gain attention and likes from netizens, and thus realize traffic monetization. Experts say that AI rumor spreading has the characteristics of customization, low threshold, difficult recognition, and great destruction, and it urgently needs to strengthen supervision. At the same time, governance can be achieved through technical means such as setting AI generated content prompt identifiers and strictly implementing network real name registration. The threshold for AI rumor making is low and the harm is great. In August of this year, a news that "Songxi County in Nanping, Fujian Province was hit by a severe strong wind, causing a death on Zaoling Street" quickly spread on the internet. This news, accompanied by vivid images and even realistic video clips, immediately attracted a large number of netizens' attention and reposting. But after investigation by the public security organs, it was found that this was actually a network rumor video fabricated by AI. The current AI network rumors have the characteristics of simulation and scale. According to different input instructions, AI can generate a large number of online rumors with different contents in a short period of time, and the generated online rumor information is more in line with human expression. Combined with deep forgery technology, it generates images, videos, and recordings, making it increasingly difficult for the public to identify the authenticity of online information Associate Professor Zhao Jingwu from the School of Law at Beihang University told reporters. The reporter randomly tested the "AI Copywriting Crash" mini program and entered content creation instructions related to the theme of "fire" in the dialog box, generating a news report that even included on-site rescue details and fire safety tips, making it difficult to distinguish between true and false. Zhao Jingwu pointed out that the harmfulness of AI network rumors mainly manifests as disrupting social public order, weakening government credibility, and inducing telecommunications network fraud. Zhao Jingwu analyzed that the motives behind individuals or organizations using AI technology to spread rumors for illegal gains are complex and diverse. This includes some netizens intentionally fabricating AI network rumors in order to seek a sense of existence and stimulation, as well as criminals fabricating AI network rumors for illegal gains in order to smoothly carry out criminal activities such as telecommunications network fraud. In mid June this year, Wang, the actual controller of an MCN organization, was administratively detained by the police for using AI software to generate false news and spreading it widely, disrupting public order. According to police reports, Wang operates 5 MCN institutions and 842 accounts. Since January this year, Wang has been generating fake news through AI software, with a peak of 4000 to 7000 articles generated per day. Recently, the Cangnan Public Security Bureau in Wenzhou, Zhejiang successfully dismantled a MCN organization called the "Internet Water Army" gang that spread rumors about old cases. Further investigation revealed that the relevant accounts and videos were used and published by a certain network technology limited company (actually a "network water army" gang operating under the MCN model). The MCN "network water army" gang collected sensitive hotspots across the network, arbitrarily set false plots, and used AI software to automatically generate multiple false copy of key plots based on the plot. They arranged for real people to appear on camera and explain, spliced misleading materials from the network, and edited them into so-called "original" videos. The gang knew that the above-mentioned videos were filled with a large amount of false and provocative information that stirred up public emotions. In order to gain traffic and illegally profit, they publicly released them on multiple platforms, which later caused a large amount of false video information to be viewed and spread, thereby seeking illegal traffic rebates. Huang Futao, a doctor of criminal law at Renmin University of China and a trainee lawyer at Beijing Jingshi Law Firm, told reporters that anyone who fabricates false danger, epidemic, disaster and police information and spreads it on the information network or other media, or knowingly spreads the false information on the information network or other media and deliberately disturbs the social order, shall be sentenced to fixed-term imprisonment of not more than three years, criminal detention or public surveillance; Those who cause serious consequences shall be sentenced to fixed-term imprisonment of not less than three years but not more than seven years. If the perpetrator not only produces and spreads false information, but also aims to defraud the property of an unspecified majority of people on the Internet platform, it is suspected of fraud. How to govern and prevent the generation of AI network rumors? In recent years, relevant departments have introduced multiple policies and measures to address the chaos of AI and deepen the governance of the network ecosystem. The Administrative Provisions on the Deep Synthesis of Internet Information Services implemented on January 10, 2023 points out that any organization or individual shall not use the deep synthesis services to produce, copy, publish, disseminate information prohibited by laws and administrative regulations, and shall not use the deep synthesis services to engage in activities prohibited by laws and administrative regulations, such as endangering national security and interests, damaging the national image, infringing social and public interests, disrupting economic and social order, infringing the legitimate rights and interests of others. The Interim Measures for the Management of Generative Artificial Intelligence Services, which will come into effect on August 15, 2023, stipulate that the provision and use of generative artificial intelligence services shall comply with laws and administrative regulations, and respect social ethics and moral standards. It is prohibited to generate content that incites subversion of state power, overthrow the socialist system, endangers national security and interests, damages national image, incites separatism, undermines national unity and social stability, promotes terrorism and extremism, promotes ethnic hatred, discrimination, violence, pornography, and false and harmful information that are prohibited by laws and administrative regulations. To strengthen judicial governance in the field of AI, it is necessary to adhere to the basic principles of 'equal emphasis on security and development' and 'graded and classified supervision'. In response to technology abuse phenomena such as generating online rumors, it is necessary to clarify the main responsibilities of developers, service providers, and online platforms in the pre event stage, and strictly pursue the criminal and administrative responsibilities of violators in the post event stage Zhao Jingwu emphasized. Meanwhile, he stated that governance can also be achieved through technological means, such as setting sensitive keywords and prohibiting the generation of significantly illegal information content; Strictly implement the network real name system and cooperate with watermark technology to achieve effective traceability of generated content; Or the platform can set AI generated content prompt identifiers, etc. (New Society)