Governing Generative AI Disinformation: A Public Interest Litigation Approach

The large-scale deployment of generative artificial intelligence has significantly lowered the cost and increased the realism of disinformation production, thereby posing systemic threats to the online information ecosystem, social order, and legally protected public interests. generative AI–generated disinformation not only amplifies traditional risks such as rumor dissemination, but also generates novel challenges through deepfakes, automated manipulation of public opinion, and large-scale infringement of collective rights. Existing regulatory approaches—primarily private litigation and administrative enforcement—have proven structurally inadequate in addressing these risks due to high enforcement costs, fragmented regulatory authority, and their predominantly ex post orientation. Against this backdrop, this article argues that prosecutorial public interest litigation, a distinctive mechanism under Chinese law, offers a viable legal pathway for governing generative AI–generated disinformation. By conceptualizing disinformation as a technology-facilitated public interest infringement, the study constructs an integrated governance framework that targets the stages of content generation, dissemination, and resulting harm. It further develops a typology of litigation models, including civil public interest litigation, administrative public interest litigation, and criminal-attached civil public interest litigation, alongside a layered liability regime that allocates responsibilities among developers, platform operators, and malicious users. Through a systematic analysis of substantive rules, procedural mechanisms, and remedial measures—such as injunctive relief, risk prevention orders, and public interest-oriented damages—this article demonstrates how judicial intervention can complement administrative regulation and enhance preventive governance. While grounded in the Chinese legal context, the proposed framework provides comparative insights for the legal governance of online information and the management of disinformation risks in the era of generative AI.