AI-generated videos spread misinformation in China and globally
AI-generated videos spread online in March 2026 falsely depicting conflict scenes between Israel and Iran, reaching millions of viewers and misleading the public. In China, a commercial ecosystem of creators used AI tools to generate synthetic videos modifying films and television content, raising concerns about copyright infringement, misinformation about historical figures, and potential harm to young people. Authorities launched investigations and content removal efforts.
What happened
AI-generated videos falsely depicting conflict scenes circulated online in March 2026. A video posted on social media on March 3 claimed to show Israeli cities under missile attack but was verified as AI-generated. Similar synthetic videos and images falsely depicted Israeli airport evacuations, Iranian military operations, and Saudi Arabian incidents.
In China, a commercial ecosystem emerged around "AI repurposing" videos that modify television and film content by changing plots, generating dialogue, and cloning actors' voices. Creators used AI software requiring only smartphones and basic tools, working 1-3 hours per minute of video content.
Reported impact
The March 3 video about Tel Aviv reached millions of views. AI-generated content made it increasingly difficult to distinguish fabricated from authentic videos. Experts noted the videos could cause public misunderstanding of historical events and posed potential harm to young people.
In China, the practice involved copyright infringement of films and TV series, violation of actors' likeness rights, and potentially distortion of heroic figures under the Heroes and Martyrs Protection Law. Video creators reported monthly earnings ranging from thousands to over 100,000 yuan.
Index analysis
Creators accessed film and television source material through cloud storage sellers on e-commerce platforms at prices from 0.99 yuan to hundreds of yuan. Some AI software required paid subscriptions after exceeding usage thresholds. A training ecosystem emerged offering year-long courses for 1,980 yuan teaching the underlying logic of AI image and video generation.
Platforms enabled the practice through reward systems and task structures available to accounts with 2,000-3,000 followers. Legal experts noted platforms as primary beneficiaries should bear responsibility for content moderation, requiring establishment of intellectual property protection systems and proactive removal of infringing material rather than waiting for complaints.
Governance controls
The Cyberspace Administration of China deployed a special action in early 2026 targeting AI-generated "digital swill" misinformation, disposing of over 39,000 accounts by the time of reporting. Legal scholars and industry experts called for platforms to establish complete intellectual property protection systems and maintain contact with rights holders.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.