The Cyberspace Administration’s annual “Qinglang” campaign arrives in a significantly different regulatory environment than last year’s edition, and the same week The White House has accused China of operating on an “industrial scale”. AI theft operations.
China has launched a month-long enforcement campaign targeting the misuse of artificial intelligence. According to Reuters.
The campaign, initiated by the Cyberspace Administration of China (CAC) and coordinated by the Ministry of Public Security and other agencies, targets AI-powered fraud, deep fraud, disinformation, and illegal applications that violate privacy and intellectual property rights.
The campaign is the 2026 edition of the “Qinglang” (Clear and Bright) special campaign series, which has become an annual enforcement mechanism. Launched on April 30, 2025, its immediate predecessor, “Correcting the Abuse of AI Technology,” lasted three months in two phases.
💜 of EU technology
The latest rumblings from the EU tech scene, a story from our wise founder Boris and some questionable AI art. Free in your inbox every week. Register now!
By the time its first phase ended in June 2025, authorities had taken down more than 3,500 AI-related products, removed more than 960,000 illegal or harmful content, and suspended or penalized more than 3,700 accounts.
This year’s campaign comes in a significantly more evolved regulatory environment and against a geopolitically charged backdrop that makes its scope and objectives more complex than its predecessor.
What does the campaign aim for?
China’s AI abuse enforcement campaigns are built around a taxonomy of abuse that expands with each iteration as both the capabilities and criminal applications of AI evolve.
Based on the Qinglang implementation framework and new regulatory measures adopted in 2025 and early 2026, this year’s campaign is expected to target multiple categories simultaneously.
The first, and most commercially significant, is AI-powered fraud and impersonation. China has seen a dramatic increase in voice cloning and the use of face-changing deepfake technology impersonating celebrities, business leaders and government officials in scams targeting ordinary consumers.
The CAC’s 2025 campaign specifically targeted the use of AI to “impersonate relatives and friends and engage in illegal activities such as online fraud” and the “illegal use of AI to resurrect the dead”, referring to the use of AI-created likenesses of people who have died without consent.
The CAC published draft rules for digital virtual human services on April 3, 2026, which include consent requirements for use of likeness and prohibit circumvention of biometric authentication systems, and the public comment window closed on May 6.
The second major category is AI-generated disinformation and “online water army” activity, the industry-wide use of AI to create fake social media accounts, create and distribute linked content, manipulate engagement metrics, and create artificial trending topics.
The 2025 campaign has identified this as a priority for its second phase, focusing on platforms that facilitate AI-powered account management, crowdsourced content creation and social bot networks.
The third is non-compliance with mandatory documentation and registration procedures. China requires large language models that offer generative AI services to the public to undergo a security assessment and complete an application to the CAC before launching.
As of March 2025, 346 generations of AI services have completed LLM documentation; there were no more. The first phase of the 2025 campaign identified undocumented AI products as a major remediation target, with local regulators, including the Shanghai CAC, penalizing three AI apps that provided services without completing the required process, and the Zhejiang CAC ordering app stores to remove a face-swapping app that had not passed a security assessment.
The fourth is the management of learning data using a learning corporation that includes content that violates intellectual property rights, privacy rights, or consent obligations.
This enforcement angle is particularly sensitive in 2026, given the White House’s official accusation on April 23 that Chinese companies are using jailbreaking techniques and tens of thousands of proxy accounts to conduct “industrial-scale” distillation campaigns to extract capabilities from US border AI models.
China’s domestic enforcement campaign does not directly address this US charge; it is aimed at protecting Chinese rights holders and users, not Americans. But the two regulatory environments now thrive with a clear understanding of each other.
The 2026 campaign works against a significantly more advanced local regulatory architecture than its predecessor. Several key regulations have either gone into effect or been published in draft form in the months leading up to this enforcement push.
China’s mandatory AIGC (AI-generated content) labeling standards, which require technical labels to appear on all AI-generated text, images, audio and video, came into effect on September 1, 2025.
On April 10, 2026, the CAC published Interim Measures for Governing Anthropomorphic AI Interactive Services, which govern chatbots, AI companions, and AI customer service agents that simulate human personality and communication styles from July 15, 2026.
On April 3, the CAC published draft rules for digital virtual human services covering biometric deep spoofing, with public comment closing on May 6, 2026. And in April 2026, the CAC, MIIT and MPS jointly published a 2026 data protection agenda targeting seven sectors.
The effect of this layered rulemaking is that the 2026 Qinglang campaign has significantly more legal teeth than its 2025 predecessor. Implementation measures can now simultaneously be based on mandatory labeling standards, LLM documentation requirements, Deep Synthesis Measures from 2023, and the new Anthropomorphic AI Intermediate Measures.
Violating companies face administrative disciplinary measures, suspension of service, forced removal within specified periods, and in severe cases, referral to state security agencies for criminal proceedings.
Geopolitical framework
The timing of the campaign gives it an unavoidable geopolitical dimension. A week before the launch, on April 23, White House OSTP director Michael Kratsios released a memo accusing China of conducting “deliberate, industry-wide campaigns to distill US cross-border AI systems” using tens of thousands of proxy accounts and jailbreaking techniques.
The charge was presented as evidence that Chinese companies, including DeepSeek in particular, had stolen US AI intellectual property by violating the terms of service of US labs.
China’s domestic AI abuse campaign this is not a response to the accusation, it is an annual structural exercise before the current geopolitical moment. But its focus on AI-powered fraud, impersonation, and data breaches implicitly reinforces the argument that both countries are grappling with the same categories of AI abuse: China is exposing its own population to AI-powered fraud and manipulation, while the US accuses Chinese actors of exploiting its own AI technologies.
The campaign also coincides with the Trump-Xi summit scheduled for May 14 in Beijing. AI and semiconductor export controls were both expected to be on the agenda, with the White House’s AI theft memo clearly complicating those talks by analysts.
China’s domestic enforcement campaign, built around protecting users and protecting social stability, is unlikely to figure directly into these diplomatic discussions, but it is part of a larger picture of how each government is positioning its domestic AI governance ahead of the high-stakes bilateral meeting.
