AI isn't just being used to spread disinformation — it's now the subject of it, and a new OpenAI report reveals a coordinated foreign influence campaign designed to shape how Americans think about AI policy.
What the Report Actually Found
OpenAI's threat intelligence team has identified PRC-linked influence operations using AI-generated content to manipulate U.S. debates around tech policy — specifically targeting narratives about data centres, tariffs, and ChatGPT itself. These aren't clumsy bot farms. They're structured, scalable disinformation pipelines.
The operations appear designed to sow distrust in American AI infrastructure and steer public opinion on regulation. In other words: the battleground for AI governance isn't just in Brussels or Washington D.C. — it's in your social feed.
Why AI Disinformation About AI Is a Different Beast
There's a sharp irony here. The same technology being debated is being weaponised to distort that very debate. AI-generated text, personas, and content make it cheaper and faster than ever to flood the information environment with coordinated narratives at scale.
This creates a compounding trust problem: as AI literacy remains low among the general public, manipulated narratives about AI capabilities, dangers, or ownership can land without scrutiny. If people can't evaluate AI claims critically, they become easy targets. Understanding how AI-generated content works — and how to spot it — is no longer optional.
If you want to understand the mechanics behind how language models produce convincing text (and why that makes disinformation so scalable), Decoding Language Models Tokenization is a solid starting point.
The Regulatory and Ethical Ripple Effects
This report lands at a critical moment. The EU is finalising its AI Act implementation, and OpenAI separately announced support for the EU Code of Practice on AI content transparency this same week. Provenance tools — metadata that tags AI-generated content at source — are being positioned as a key defence layer.
For businesses deploying AI, this raises the stakes on governance. Being caught amplifying — even accidentally — manipulated AI narratives carries serious reputational and regulatory risk. Leaders building AI strategy need to factor information integrity into their frameworks, not just model performance. Our AI Strategy for Senior Leaders course covers exactly this kind of second-order risk thinking.
And for anyone worried about the darker failure modes of AI systems in adversarial conditions, When AI Goes Rogue explores how AI can be misused, manipulated, or turned against its intended purpose.
What This Means for Learners
The most important AI skill right now might not be prompting — it might be critical evaluation. Being able to identify AI-generated content, understand how influence operations work, and assess AI claims with scepticism is becoming a core digital literacy requirement.
The people best equipped to resist AI-powered disinformation are the ones who understand how the technology actually works. That's not paranoia — that's the new baseline for informed citizenship in an AI-saturated information environment.
