- A NewsGuard study revealed that a Moscow-based disinformation network, “Pravda,” has infiltrated Western AI chatbots, causing them to spread Russian propaganda by flooding the web with 3.6 million pro-Kremlin articles in 2024 alone.
- The audit tested 10 major AI chatbots, including ChatGPT, Grok, and Gemini, finding they repeated Pravda’s false narratives 33 percent of the time, with seven directly citing Pravda sites as credible sources.
- Pravda employs “LLM grooming,” manipulating AI training datasets with disinformation, such as false claims about Ukraine’s President Zelensky banning Trump’s Truth Social app, which six chatbots echoed.
- Researchers warn this subtle manipulation, amplified by AI’s scale, poses a widespread threat to information integrity, as chatbots fail to distinguish propaganda from fact. Read More
Menu
