AI-generated uniform photos used for fraud and impersonation in China
Criminals in China have used AI costume generation tools to create deepfake photographs and videos of themselves in military and police uniforms to commit fraud and impersonation, exploiting inadequate platform content review and a lack of synthetic content labeling to deceive victims.
What happened
Users of AI costume generation tools can upload a photograph and generate images of themselves wearing military or police uniforms within seconds. Reporters found multiple AI applications on short video platforms offering this feature, marketed with slogans such as 'wear an electronic military uniform for fun' and 'AI realized my childhood dream.'
Criminals have misused these tools to commit fraud. In April 2025, Lu (路某) used AI to synthesize his own face into military uniform photos, posted stolen photos of military personnel on short video platforms, fabricated a military officer certificate, and fraudulently posed as a Major to deceive six women, defrauding them of over 30,000 yuan. The Jiangyin People's Court sentenced him to two years imprisonment for impersonating a military officer and swindling. In March 2025, Jiang (江某) from Anhui's Huangshan city used AI to create military uniform photos and falsely claimed to have been appointed Commander of Huangshan Military District by military order, attracting online attention. In July 2025, scammers in Changzhou used AI-generated video of a person in police uniform to intimidate an elderly resident into paying a 'guarantee fee' for alleged involvement in a case, until police intervention prevented financial loss.
Reported impact
The technology lowers the technical threshold for identity forgery. AI-generated military and police uniform images can be purchased from online merchants for less than one yuan, with template files available for photoshop and AI manipulation. Platforms lack meaningful review mechanisms when users replace profile pictures with AI-generated uniform images.
Experts expressed concern that proliferation of non-compliant military and police images—mixing military branch uniforms, using fake medals, or combining elements from different countries' militaries—undermines public trust and respect for military and police personnel.
Index analysis
Content platforms and third-party software lack adequate review mechanisms to detect and prevent AI-generated uniform images, despite China's September 1, 2025 regulation requiring AI-synthesized content to be clearly labeled. Researchers found that AI costume generation applications do not request identity verification, do not warn users against illegal use, and do not label generated content as synthetic.
Governance controls
China's 'Artificial Intelligence Generated and Synthesized Content Marking Measures' (effective September 1, 2025) requires AI-synthesized content to be clearly labeled. The 'Internet Military Information Dissemination Management Measures' (issued by the Cyberspace Administration and nine other departments in 2025) prohibits the use of deepfakes and generative AI to damage the image of the People's Liberation Army.
The Cyberspace Administration of China has taken enforcement action against a Zhejiang company operating an app that provided deepfake capabilities without proper safety assessments or content labeling, ordering the app's removal from distribution platforms for violating deepfake and generative AI regulations.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.