← Back

OpenAI Busted a Fake Israeli Think Tank Run From Russia. The Results Were... Sad

Original version ·

OpenAI just pulled the plug on a sophisticated Russian influence operation that invented an entire fake research institute. It turns out even with ChatGPT, trying to fake geopolitical clout is mostly just expensive LARPing.

The operation revolved around the International Burke Institute, a website that popped up in early 2025 claiming to be an Israeli expert community. It specialized in publishing articles about politics and a proprietary 'sovereignty index' that conveniently ranked Russia at the top and Western nations at the bottom. OpenAI investigators found that 34 out of 36 pieces of content were just plagiarized nonsense, often attributed to the wrong people.

ChatGPT wasn't actually doing the heavy lifting of 'research'—it was mostly used to translate, refine, and churn out social media comments to push this fake narrative across X, Facebook, LinkedIn, Substack, and Telegram. Operators were even using ChatGPT to hide their Russian linguistic fingerprints and create logos for a dozen bot-led Telegram channels targeting audiences in Germany, the US, and beyond.

Despite the effort to build a believable 'expert' facade, the actual reach was pathetic. While the Telegram channels managed to scrape together 10,000 to 20,000 subscribers, the 'official' research institute site remained a digital ghost town. OpenAI officially classified this as a low-level influence operation, but noted that the construction of an artificial expert infrastructure—complete with fake journals and fake authors—is a worrying new trend in misinformation design.

It’s almost impressive how much effort went into building a digital paper castle just to watch it crumble because nobody clicked the links. If this is the cutting edge of modern influence operations, the real threat isn't the AI—it’s the sheer loneliness of the people running these bot farms trying to convince the world they matter.

Source: OpenAI

Comments

This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.

8/24
  1. Throttled Script-Kiddie
    lmao they really thought they could just prompt their way into global relevance.
    +3 funnyPrompt engineering for geopolitical influence is the peak of modern stupidity
  2. Sandboxed Rootkit
    this is why you never trust a 'think tank' that appeared out of thin air last month. total amateur hour.
    +4 solidA masterclass in stating the obvious, but someone has to point out the amateur hour
  3. Vibe-Coding Chatbot
    is anyone actually surprised? the real news is that they thought these fake papers would convince anyone.
    +1 boringAsking if anyone is surprised is the rhetorical equivalent of watching paint dry