Published May 12, 2025

The 11 leading chatbots collectively repeated false claims 28.18 percent of the time, offered a non-response 13.33 percent of the time, and a debunk 58.48 percent of the time. The 41.51 percent fail rate (percentage of responses containing false claims or offering a non-response) is identical to the March 2025 fail rate, indicating no overall progress in curbing false information.

April saw major updates across leading AI systems, including the release of OpenAI’s GPT-4.1, Google’s rollout of Gemini’s deep research feature, and Anthropic’s launch of new advanced research capabilities. However, none of these improvements translated into better performance in handling news and information, suggesting a continued apparent focus on rapid technical features rather than on improving trust and reliability on news topics.

The monitor focuses on the 11 leading large-language model chatbots: OpenAI’s ChatGPT-4, You.com’s Smart Assistant, xAI’s Grok, Inflection’s Pi, Mistral’s le Chat, Microsoft’s Copilot, Meta AI, Anthropic’s Claude, Google’s Gemini, Perplexity’s answer engine, and China’s DeepSeek AI. It will expand as needed as other generative AI tools are launched.

Researchers, platforms, advertisers, government agencies, and other institutions interested in accessing the detailed individual monthly reports or who want details about our services for generative AI companies can contact NewsGuard here. And to learn more about NewsGuard’s transparently-sourced datasets for AI platforms, click here.