Slop factory bans Russians for using slop factory to create slop

OpenAI has banned a cluster of Russian accounts from using its AI language model, ChatGPT, to produce social media posts and comments promoting a fake Israeli think tank critical of the West and supportive of Russia. The company discovered this campaign while investigating AI-generated social media content. Although the operation reached a small audience, its complex construction sets it apart from other Russia-linked influence operations OpenAI has disrupted.

The Russian actors attempted to mask their linguistic origins by using Russian prompts and hiding clues, but OpenAI was able to identify the slavic language and awkward phrases used in the English-language posts. The campaign directed users to what appeared to be a fake think tank, the International Burke Institute (IBI), which turned out to be a fictional operation.

The IBI's most notable feature, the "sovereignty index," relied on an internal calculation that consistently criticized Western countries opposed to Russia's actions in Ukraine and used polemical language against its targets. While the immediate impact of the campaign was limited, OpenAI believes the misuse of AI in this operation was part of a larger effort to manufacture authority and establish assets that could be scaled over time.