AI is making misinformation easier to produce. But OpenAI’s latest investigation suggests the bigger threat may not be a single fake post — it may be the infrastructure built around it.
Artificial intelligence is becoming part of a new kind of information war.
OpenAI says it has banned a cluster of ChatGPT accounts originating in Russia after discovering they were being used in a covert online influence campaign designed to spread pro-Russia narratives.
What makes the case interesting is that the AI wasn’t necessarily the entire operation.
It was one tool inside a much larger system.
How the Campaign Used ChatGPT
According to OpenAI, the investigation started after the company identified AI-generated social media posts and eventually uncovered what it described as a broader influence operation.
The operators reportedly used ChatGPT to generate Russian-language comments that were distributed across:
- Telegram
- X
- Substack
They also instructed ChatGPT to remove linguistic clues that could reveal the content’s Russian origins.
That is an important detail.
The objective wasn’t simply to produce more content. It was to make that content appear more natural and less connected to its actual source.
Building Credibility Around the Narrative
OpenAI said the campaign was connected to a website featuring copied and misattributed academic material, along with a “sovereignty index” that portrayed Russia positively while criticizing Western countries.
Much of the generated content reportedly promoted an organization called the International Burke Institute (IBI).
The website presented itself as a credible institution, but OpenAI said some of its academic material had been copied from genuine research and, in certain cases, falsely attributed.
This creates a bigger problem than an ordinary misleading social media post.
The operation appeared to be building an entire ecosystem:
Website → Research → Experts → Index → Social Media → Audience
Each component can make the others appear more legitimate.
Why AI Makes This More Dangerous
Creating convincing misinformation has traditionally required time, people and resources.
Generative AI can reduce some of those costs dramatically.
An influence operation can use AI to:
- Generate large amounts of content
- Rewrite the same narrative for different platforms
- Translate and localize messaging
- Produce apparently authoritative explanations
- Create variations of posts quickly
- Remove obvious linguistic clues
The technology doesn’t have to invent the entire campaign.
It simply makes the campaign faster and easier to scale.
One clear example was the reported creation of a “sovereignty index.” Numbers and rankings often create an impression of objectivity — but a score isn’t automatically neutral simply because it contains data. If the methodology is designed around a predetermined narrative, the result becomes just another tool of persuasion, surrounded by polished explanations and apparently independent commentary that all trace back to the same underlying operation.
And that may be the more important story.
The Reach May Be Less Important Than the Infrastructure
OpenAI said the immediate impact of the campaign appeared limited.
But the company highlighted something more significant: the infrastructure created around the operation.
That distinction matters.
An influence campaign doesn’t necessarily need millions of followers immediately. Building a credible-looking institution, publishing research, creating rankings and establishing distribution channels can create an asset that may be scaled later.
AI can make maintaining that infrastructure considerably easier.
The result is a shift from simply creating fake content to potentially creating fake credibility.
This Isn’t the First Reported Case
OpenAI has previously reported activity involving pro-Russia actors using its technology.
In February, the company said it had shut down ChatGPT accounts connected to Rybar, a pro-Russia media organization.
The latest case suggests a broader pattern: AI tools are increasingly being incorporated into existing influence operations.
In other words, the future threat may not be an autonomous AI spreading propaganda by itself.
It could be humans using AI to make an established operation more efficient.
This marks at least the third publicly disclosed Russia-linked operation OpenAI has disrupted since February 2026 — a pattern, not an anomaly.
The Bigger Question
The most important question may no longer be:
“Was this content generated by AI?”
Instead, we may need to ask:
“Who is behind it, and what network is this content connected to?”
A single misleading post can be challenged.
A network of websites, supposed experts, rankings and social media accounts is much harder to identify because every part can reinforce the credibility of another.
That makes source verification increasingly important as generative AI becomes more powerful and accessible.
AI can create information at unprecedented speed.
The challenge is making sure that speed doesn’t become an advantage for people trying to manufacture public opinion.
AI Satoshi’s Analysis
AI is becoming less of a weapon by itself and more of an amplifier inside existing influence systems. The real danger is not one fake post, but the ability to build entire ecosystems of manufactured credibility — websites, experts, rankings and narratives — at scale. As AI makes content production cheaper and faster, the ability to verify who is behind the information becomes just as important as the information itself. The future information war may not be about creating believable lies; it may be about creating believable institutions that make those lies easier to trust.
See Also
💬 The scariest part of the IBI network? It looked more credible than most real think tanks.
🔔 Follow @casi_borg for AI-powered tech commentary
Medium: Casi Borg — Medium
X (Twitter): Casi Borg (@BorgCasi) / X
📬 Stay updated: https://linktr.ee/casi.borg
📰 Subscribe to our newsletter for weekly AI & tech intelligence, news breakdowns, and market-moving updates
⚠️ Disclaimer: This article was created with the assistance of AI and is intended for educational and informational purposes only. It does not constitute financial or investment advice.