According to a new report, Russian propaganda may be affecting certain answers from AI chatbots, such as Openai’s ChatGpt and Meta’s Meta AI.
NewsGuard is a company that develops rating systems for news and information websites, claiming it has found evidence that a Moscow-based network named “Pravda” has published false claims to affect AI model responses.
Pravda has published misleading articles of 3.6 million in 2024 alone with search results and web crawlers flooded with falsehoods from pro-Lucia, and cites statistics from the nonprofit Sunlight project, according to News Guard.
A NewsGuard analysis, which examined 10 major chatbots, found that chatbots reiterated the stories of Russian disinformation in bulk.
According to NewsGuard, the effectiveness of Pravda networks in penetrating AI chatbot outputs can be largely attributed to its technique, and involves search engine optimization strategies to increase content visibility. This could prove to be a refractory issue for chatbots that rely heavily on web engines.
Source link
