Community Trust ScoreVerified
A Russian-linked network used ChatGPT to fabricate an entire academic institution. OpenAI’s report, released Wednesday, laid it all out — fake profiles, plagiarized research, and a carefully constructed lie dressed up as legitimate scholarship.
OpenAI banned multiple accounts tied to the operation. The accounts used Russian-language prompts to generate English social media content, then masked the content’s Russian origins through ChatGPT. The audience the campaign actually reached was pretty limited, but its complexity set it apart from other Russia-linked influence operations disrupted since the Ukraine conflict began. That gap between reach and sophistication is what caught investigators’ attention.
The International Burke Institute Fraud
The fake organization at the center of all this was called the International Burke Institute. It claimed to be an Israel-based expert group. Official-looking profiles and accounts mimicked ordinary users to make the institute seem credible. ChatGPT generated content that appeared authoritative enough to pass a casual glance.
It didn’t survive a closer one.
OpenAI’s review of the institute’s website found that 34 out of 36 articles were plagiarized. Some of those articles credited academics who weren’t the original authors. One example: a piece on the China-Pakistan Economic Corridor, lifted directly from a Cambridge University Press source, was misattributed to a University of Nottingham professor who specializes in South Asian politics. That professor probably had no idea their name was attached to stolen work.
Because ChatGPT isn’t available in Russia, the group used virtual private networks — VPNs — to access the service. The VPNs let them bypass regional restrictions and generate content that appeared to originate from somewhere else entirely. Smart move, operationally. It bought them time before anyone started looking hard at the institute’s actual output.
Doppelganger and Forest Blizzard: Not a New Pattern
The International Burke Institute operation didn’t happen in a vacuum. In 2024, OpenAI exposed a separate Russia-linked campaign called Doppelganger. That one used ChatGPT to create content favoring Russia while criticizing Ukraine, NATO, and the United States. Same basic toolkit, different target.
And it goes further than influence campaigns. Groups linked to Russian military intelligence have also used ChatGPT for cyber operations. A group called Forest Blizzard — connected to Russian military intelligence — exploited the chatbot for target research, scriptwriting, and phishing support. That’s not just disinformation. That’s AI being used as a direct operational tool in cyber warfare.
The range of applications is what makes it concerning. Fabricating academic institutions. Shaping geopolitical narratives. Building phishing infrastructure. It’s the same technology doing very different jobs, all pointing in the same direction.
What OpenAI Still Can’t Answer
Despite all of this, OpenAI didn’t attribute the International Burke Institute campaign directly to the Russian government. The language in the report was careful. The use of Russian-language prompts doesn’t automatically equal state direction, and OpenAI said it can’t draw that line with confidence.
Some individuals in Israel may have represented the institute at conferences or contributed articles to it. Their connection to the actual campaign, though, remains undefined. OpenAI said it’s unable to determine any direct links between those individuals and the Russian operators running the operation. Unclear whether they were witting participants, unwitting ones, or something messier in between.
That ambiguity is genuinely frustrating from an accountability standpoint. The network went to serious lengths — fake personas, stolen academic content, VPN access — and yet the chain of command stays murky. Decentralized operations are harder to trace. That’s probably the point.
What the report does make clear is that the sophistication of these campaigns keeps climbing. Generative AI lowers the cost of producing convincing content dramatically. A network that might have needed a team of writers to maintain a fake academic website can now do it faster and cheaper with a chatbot. The plagiarism problem — 34 of 36 articles stolen — suggests the operators weren’t even trying that hard on quality. They were betting on volume and surface credibility.
It’s worth noting that the International Burke Institute campaign, despite its complexity, didn’t actually reach a wide audience. OpenAI flagged that explicitly. But reach isn’t always the goal in early-stage influence operations. Sometimes it’s about building infrastructure — profiles, websites, article archives — that can be activated or scaled later.
Forest Blizzard used ChatGPT for phishing support. The International Burke Institute used it for academic impersonation. Doppelganger used it for geopolitical narrative shaping. OpenAI has now banned accounts tied to all three, but the broader pattern of AI-assisted influence operations tied to Russia-linked actors keeps expanding.
OpenAI said 34 of the 36 articles on the institute’s site were plagiarized.
Frequently Asked Questions
What was the International Burke Institute?
The International Burke Institute was a fabricated organization promoted by a Russia-linked network using ChatGPT to pose as an Israel-based academic expert group, with 34 of its 36 website articles found to be plagiarized.
How did the network access ChatGPT if it’s unavailable in Russia?
The operators used virtual private networks (VPNs) to bypass regional restrictions and access ChatGPT, allowing them to generate content that appeared to come from outside Russia.
Is the campaign linked to the Russian government?
OpenAI did not attribute the International Burke Institute campaign directly to the Russian government, saying it was unable to determine any direct links between the operators and state entities.
Why It Matters
The emergence of a fake academic institution leveraging advanced AI tools like ChatGPT highlights the increasing sophistication of disinformation campaigns, particularly those linked to state actors. This incident underscores the challenges facing regulatory bodies and social media platforms in combating misinformation, as well as the potential implications for academic integrity and trust in research. As AI technology continues to evolve, the risks associated with its misuse for propaganda and manipulation could pose significant threats to information ecosystems and public discourse globally.
