A leak of files from the laptop of former U.S. Marine John Mark Dougan, who became a Russian propagandist, revealed the scale of his disinformation work. The documents show that he systematically used artificial intelligence models — in particular Claude from Anthropic and Qwen from Alibaba — for the mass generation of pro-Russian articles and manipulation of other chatbots.

How the AI-based content "factory" worked

Data about Dugan's working methods ended up online after unknown individuals hacked his account on the social network X and published a link to a twin website with his emails, correspondence in Telegram and Signal, as well as a history of queries (prompts) to AI, writes Cybernews. Dugan himself, in a comment to NewsGuard, confirmed the authenticity of the documents, stating that he does not care about the publication of the contents of his laptop.

According to the leak, the main goal of the campaign was the so-called "infection" of large language models (LLMs) and search algorithms:

  • the network was creating up to 43,000 articles per month that praised russia and masqueraded as legitimate materials for a Western audience.

  • articles were constantly updated with AI-generated text to improve their search engine rankings and make them available as "sources" for other chatbots;

  • the expenses for this activity amounted to about 2 million rubles (approximately $23,000) per month. In one of the prompts to Claude, Dugan asked the AI to draw up an estimate, complaining that the current volume of 43,000 texts was becoming too expensive.

    • Bypassing Anthropic and Alibaba security policies

    For his propaganda work, he used Anthropic Claude, the Chinese model Alibaba Qwen, as well as the Open WebUI interface, which allows running models on one's own terms. Such activity directly violates the terms of use of both platforms, which prohibit the generation of fakes and disinformation.

    In correspondence with Claude, Dugan, in particular, asked it to develop a concept for a "free speech website for Europe" aimed at a French-speaking audience, consulting with the bot about the name. And he approached the Qwen model with requests on how to write texts so that they could not be detected by AI-content detectors.

    Who is John Mark Dougan

    John Mark Dougan (also known by the pseudonym BadVolf) previously served in the U.S. Marine Corps and worked as a deputy sheriff in the states of Florida and Maine, from where he was dismissed for misconduct and harassment.

    In 2016, the FBI launched an investigation against him over the leak of confidential police records. After 21 criminal charges were filed in Florida, Dougan fled to russia, where he received political asylum, and in 2023 — russian citizenship. According to investigators, he collaborates with russian military intelligence (GRU), which funds his network of fake websites disguised as American media.