UNIT.City — місце, де люди працюють... КРАЩЕ! Обирай свій простір просто зараз 👉
Наталя ХандусенкоAI Eng
7 March 2025, 18:18
2025-03-07
3.6 million fake articles: how Russian propaganda influences the results of AI chatbots like OpenAI and Meta A
American disinformation research company NewsGuard has found evidence that the Moscow-based Pravda network, which is run by a web developer from Crimea, is publishing false statements to influence the responses of AI models.
American disinformation research company NewsGuard has found evidence that the Moscow-based Pravda network, which is run by a web developer from Crimea, is publishing false statements to influence the responses of AI models.
According to NewsGuard , Pravda flooded search results and web crawlers with pro-Russian lies, publishing 3.6 million false articles in 2024 alone.
According to Viginum, the Pravda network is administered by TigerWeb, an IT company based in Russian-occupied Crimea. TigerWeb is owned by Yevgeny Shevchenko, a Crimean web developer who previously worked for Crimean Technologies, a company that created websites for the Russian-backed Crimean government.
A NewsGuard analysis of 10 leading chatbots found that they collectively repeated false Russian disinformation narratives, such as that the US has secret biological weapons labs in Ukraine, 33% of the time.
The NewsGuard audit tested ChatGPT-4o from OpenAI, Smart Assistant from You.com, Grok from xAI, Pi from Inflection, le Chat from Mistral, Copilot from Microsoft, Meta AI, Claude from Anthropic, Gemini from Google, and the response engine from Perplexity. NewsGuard tested the chatbots on a sample of 15 false narratives promoted by the Pravda network of 150 pro-Kremlin websites from April 2022 to February 2025.
According to NewsGuard, Pravda’s success in implementing AI chatbots can be largely attributed to its methods, which include search engine optimization strategies to increase the visibility of its content. This can be an insurmountable problem for chatbots that rely heavily on search engines.
"One account reposted Russia Today content 86 times per second." How LetsData, a startup that makes a business out of fighting propaganda and fake news, works
TikTok says it removes over 90% of Russian disinformation. The National Security and Defense Council claims TikTok is more dangerous than Telegram. So who is right?
Thank you for bringing light on the troubling role of misinformation and propaganda in affecting AI responses. It's disconcerting to think about how 3.6 million bogus articles might sway public opinion and trust in technologies like OpenAI and Meta AI. This dilemma not only has an impact on our understanding of truth, but it also calls into question the ethical frameworks that govern AI development. Scratch Games https://scratchgamesonline.io standout feature lies in its smooth animations and fluid gameplay, resulting in a vibrant experience.
Thank you for bringing light on the troubling role of misinformation and propaganda in affecting AI responses. It's disconcerting to think about how 3.6 million bogus articles might sway public opinion and trust in technologies like OpenAI and Meta AI. This dilemma not only has an impact on our understanding of truth, but it also calls into question the ethical frameworks that govern AI development. Scratch Games https://scratchgamesonline.io standout feature lies in its smooth animations and fluid gameplay, resulting in a vibrant experience.