OpenAI has exposed and banned ChatGPT accounts linked to two covert influence operations: Russia's 'Dark Clark' and Iran's 'Bogus Bylines.' Rather than relying on social media, both campaigns used fake personas to plant AI-generated articles in legitimate news outlets. 'Dark Clark' targeted Latin America, triggering official denials in Ecuador and Peru, while 'Bogus Bylines' placed nearly 100 articles about the U.S.-Iran conflict in publications like Daily Kos and the Port St. Joe Star.
OpenAI investigation reports
- ▪According to an OpenAI investigation report, OpenAI exposed and banned ChatGPT accounts associated with the Russian 'Dark Clark' and Iranian 'Bogus Bylines' influence operations, which used ChatGPT to generate content
- ▪OpenAI has been reporting on influence operations for two and a half years
Common tactics of both operations
- ▪Both the Russian 'Dark Clark' and Iranian 'Bogus Bylines' operations used artificial intelligence primarily for internal reporting and to adapt propaganda into different languages
- ▪The Russian 'Dark Clark' and Iranian 'Bogus Bylines' influence operations exposed by OpenAI used fake identities to plant content in legitimate media outlets rather than running traditional social media campaigns
The Russian Dark Clark campaign
- ▪The Russian influence operation, named "Dark Clark," spread disinformation across Latin America to discredit Ukraine and destabilize local politics
- ▪Fabricated audio files and documents from the "Dark Clark" operation triggered fact-checks and official denials from politicians in Ecuador and Peru
- ▪The "Dark Clark" operation controlled a think tank through a fictitious persona and likely hired local staff members who were unaware of the operation's true nature
- ▪OpenAI rated the "Dark Clark" campaign as a category 5 out of 6 on its Breakout Scale due to reactions from politicians, marking the first category 5 case in two and a half years
Debatable claims
- ▪AI developers bear the primary responsibility for policing foreign influence operations on their platforms
- ▪Publicly reporting on foreign cyber threats is an appropriate role for private AI companies
- ▪Generative AI has become the primary threat in foreign disinformation campaigns
- ▪The political impact of AI-generated foreign disinformation is widely exaggerated
Story comments
Loading comments…