Russian propaganda can influence certain responses from AI chatbots, including the Openai and Meta’s Meta AI, According to a new report.
NewsguardA company that develops rating systems for news and information websites, claims to have found evidence that a network based in Moscow named “Pravda” publishes false claims to allocate the responses of AI models.
Pravda flooded research results and web robots with pro-Russian lies, publishing 3.6 million deceptive articles in 2024, by Newsguard, citing statistics from the non-profit organization US sunlight project.
The analysis of Newsguard, which surveyed 10 leading chatbots, revealed that chatbots have collectively repeated the false accounts of Russian disinformation, such as the United States, the United States exploiting bio-armes laboratories in Ukraine, 33% of the time.
According to Newsguard, the efficiency of the Pravda network in the infiltration of AI chatbot outputs can be largely attributed to its techniques, which involve strategies for optimizing search engines to stimulate the visibility of its content. This can be an insoluble problem for chatbots highly dependent on web engines.
