OpenAI has banned a Russia-origin influence operation that used its models to churn out propaganda.
The company identified a network it calls Stop News generating multilingual articles and social posts aimed at audiences in Ukraine and across the West. The accounts used OpenAI's tools to draft and translate content at a pace and volume no human newsroom could match. OpenAI removed the accounts as part of its ongoing effort to disrupt malicious use of its models. The company has not disclosed how many accounts were involved or how widely the material spread before it was caught.
This is the same playbook influence operations have used for years - translated, mass-produced content designed to look native - but AI removes the old bottleneck of hiring writers in multiple languages. That makes disinformation campaigns cheaper to run and harder to spot at a glance. It's also a reminder that AI labs like OpenAI are now, whether they asked for it or not, part of the internet's content-moderation infrastructure.
Banning the accounts stops this operation from using OpenAI's tools. It does nothing to stop the same group from switching to another model, or none at all.