AI made established tactics easier to scale
OpenAI says it banned two influence operations, one linked to Russia and one to Iran. The Iranian operation used seven “journalist” personas to pitch long-form stories to smaller outlets. The Russian operation appeared to involve unwitting people in Latin America operating a supposed think tank.
The report says the operations also generated comment batches, fake leaked documents and audio scripts. OpenAI describes AI as an accelerant for existing influence techniques rather than a wholly new playbook. The company says its review found that some of the content was not generated by its models.
The route to an audience matters
OpenAI rated the Russia-origin operation Category 5 on the Brookings Breakout Scale and the Iran-origin operation Category 4. It says both had content land in mainstream media outlets, a route it sees as more consequential than fake social distribution alone.
These are findings from the platform provider’s own threat research. The practical takeaway is to inspect the provenance of plausible-looking articles and “leaked” files, especially when multiple sources trace back to a network of new personas. AI can make content production cheaper; editorial verification remains the barrier between a fabricated story and a credible one.
Source published 2026-10-08. Coverage is based on the maker’s announcement and demonstration.