OpenAI disrupts two sophisticated Russian and Iranian campaigns that weaponized ChatGPT
OpenAI said it halted two covert influence operations, “Dark Clark” and “Bogus Bylines,” that used ChatGPT to spread disinformation targeting Ukraine and the United States.
On Thursday, OpenAI disclosed that it had intervened in two covert influence operations named “Dark Clark” and “Bogus Bylines,” both of which leveraged the ChatGPT AI system to generate disinformation. The Russian-linked “Dark Clark” campaign produced fake activity reports, counterfeit confidential letters, bogus contracts and fabricated audio clips, then disseminated them broadly on social networks, achieving a virality rating of 5 out of 6.
The Iranian-linked “Bogus Bylines” operation created false articles that were subsequently picked up by legitimate media outlets. Each campaign sought to undermine the image of Ukraine and the United States, and OpenAI described them as the most dangerous influence attempts it has encountered, reaching a large audience before being disrupted.
Why it matters
It reveals how AI tools can be misused by foreign actors to spread false narratives that influence public opinion and international relations.
In this story
