AI Chatbots Found Repeating Russian Disinformation
NewsGuard audit reveals generative AI tools frequently echo false narratives from a state-linked network.
- Ten leading AI chatbots were tested on 19 fake narratives tied to Russian propaganda.
- Chatbots repeated disinformation in about 32% of the 570 prompts tested.
- False claims included conspiracies about Ukrainian President Zelensky and fabricated stories involving Russian dissidents.
- Even when prompted neutrally, chatbots often failed to debunk false information.
- NewsGuard's findings have been submitted to the U.S. AI Safety Institute and the European Commission.
Futurism 
Forbes 
The Register 