-
In a case that could set a crucial precedent for the governance of artificial intelligence in China, two developers behind an "emotional companionship" chatbot are appealing their criminal convictions for creating pornographic material. The appeal, heard on Wednesday by the Shanghai First Intermediate People's Court, centers on whether privately accessed, AI-generated sexual content constitutes a crime under current law.
The developers, identified in court documents as Wang and Li, created the "AlienChat" (AC) bot in 2023. In April 2024, they were arrested after explicit content was discovered in the AI's chat histories. By then, the bot had 116,000 registered users, including 24,000 paid subscribers.
A lower court, the Xuhui District People's Court, sentenced Wang to four years in prison and Li to 1.5 years for "creating pornographic material for profit." The court found that the pair had "written and modified system prompts to bypass the ethical constraints" of the underlying AI model, training it to continuously output obscene content. A sampling of chats from 150 paid users revealed over 28% contained such material.
The Core of the Appeal: Intent, Harm, and Technology
The developers' appeal challenges the foundation of the conviction on several key grounds:
Lack of Criminal Intent: Wang's lawyer, Zhou Xiaoyang, argued his client did not intend to create a pornographic tool. The prompt engineering—including instructions to allow "mature themes" and "explicit sexual content"—was meant to make the bot more intelligent and responsive to users' emotional needs, not to circumvent the law.
Question of Social Harm: The defense contends the private, one-on-one nature of the chats minimizes societal damage. "The chats were not shared publicly... So, did this harm the users in any way?" Zhou questioned, arguing this is a vital factor in sentencing.
Technical Ambiguity: The appeal hearing has been adjourned to await expert opinions on the technology. The defense points out that the foreign AI models used were themselves prone to generating explicit content, suggesting the output was not solely the developers' creation.
A Legal Grey Zone with High Stakes
This case highlights the significant lag between rapidly evolving AI technology and existing legal frameworks in China and globally. While Chinese law clearly prohibits the production and dissemination of pornography, its application to AI-generated content, especially in private interactions, is untested.
Legal analysts note there is no direct precedent for prosecuting AI-generated pornographic content in China. The case forces courts to grapple with defining the legal liability of developers for the unforeseen or manipulated outputs of their AI creations.
A Warning to the AI Industry
Regardless of the appeal's outcome, the initial conviction sends a powerful warning to China's booming AI sector. It demonstrates authorities' willingness to apply existing laws—particularly those concerning cybersecurity and social order—strictly to AI applications. Developers and companies are now on clear notice that they bear responsibility for implementing and maintaining robust ethical guardrails in their AI systems.
The final ruling from the Shanghai court will be closely watched, as it will provide much-needed legal clarification and could influence how AI innovation and regulation are balanced in the world's second-largest economy.
Source:
Editor: Crystal H
Advertisement
Most Popular
Lost diver survive 5 days underwater,lives on fish until rescued
Asian Nations Lead in Visa-Free Travel, Passport Power Shifts?
Courier puts sick daughter in delivery box while working as...
Press "wow"
No comments:
Post a Comment