France’s Viginum, the government agency tasked with countering foreign digital threats, has exposed a network of 13 counterfeit news websites spreading pro-China propaganda in various languages. The websites are directly tied to China Global Television Network (CGTN), the state-controlled broadcaster under the Chinese Communist Party. The operation, codenamed “Fawn Mianju” by French services was announced on 4 June 2026.
The fake sites, launched primarily in 2025, posed as independent news platforms in multiple languages, including French (“Actu Méridien”), English, Spanish (“Amigo News”), and Vietnamese. They systematically promoted a one-sided, pro-Beijing narrative aimed at influencing public opinion, particularly among younger audiences in Western countries and French-speaking regions of Africa. The content heavily emphasised China’s leadership in aviation and artificial intelligence, its role as a champion of the Global South, its commitments to ecological transition, and the purported advantages for nations like France in aligning with Chinese interests. One article across the network criticised a France 2 television report on the treatment of Uyghurs in China.
These pseudo-news outlets operated through sophisticated automation. Investigators found that out of nearly 3,000 articles published by CGTN since late 2025, more than 2,300 were quickly repurposed, often within an hour, and lightly rephrased for the fake network. The sites employed large language models (LLMs) and AI tools to adapt and generate youth-oriented versions of the original CGTN material. Stylometric analysis revealed a clear reduction in the natural variation of sentence lengths and punctuation typical of human writing, providing strong indicators of semi-automated AI involvement.
Technical traces also strengthened the investigation. The domains were registered in Beijing and hosted on Alibaba Cloud. The sites used a distributed architecture across multiple servers and paid SEO plugins, signalling substantial resources. Crucially, one administrator of “Actu Méridien” and associated platforms inadvertently left traces of his login credentials, leading investigators to identify him as a senior project manager in CGTN Digital, the network’s digital arm. His public GitHub profile revealed extensive work on LLM integration for automated content generation, including digital keys for directly connecting websites to AI systems.
The network, which also included coordinated social media accounts on platforms such as Facebook and Threads, ran advertising campaigns targeting users in as many as 89 countries. Despite these efforts, its reach remained limited, with individual reports getting no more than around 15,000 views. Many interactions appeared inauthentic, concentrated in places like Burundi. French analysts said that the overall effort was an operational failure, noting that several sites, including “Actu Méridien,” have been inactive for months.
This case comes after an earlier exposé by the American cybersecurity firm Graphika in summer 2025, which had flagged 11 of the sites and related accounts. Viginum’s deeper investigation, spanning several months, has provided conclusive evidence of direct coordination with CGTN.
It shows Beijing’s willingness to pursue covert influence operations in Western and African information spaces, its advancing use of AI for propaganda, and its focus on youth audiences.
