Leading AI Chatbots Fail to Detect AI-Generated Videos in New Test

• January 24, 2026

Leading AI chatbot platforms have demonstrated significant shortcomings in identifying AI-generated video content, according to recent tests. Researchers evaluated the ability of major chatbots to detect videos created by OpenAI’s Sora generator, revealing that even ChatGPT struggled to determine whether videos were AI-produced or genuine.

The assessment was conducted by specialists at NewsGuard who loaded videos generated by OpenAI’s Sora into several popular AI chatbots. Among the tested models were ChatGPT, Gemini, and Grok, all of which were unable to accurately recognize the synthetic nature of the videos. This outcome highlights challenges in current AI systems for distinguishing between authentic and artificially generated digital media.

While these AI chatbots excel at natural language processing and text-based tasks, their performance in analyzing and verifying video content appears limited. The test results indicate potential vulnerabilities in AI-assisted content moderation and verification processes, as reliance on these models to flag AI-generated videos could lead to oversight in identifying synthetic media.

Implications for AI Content Recognition

This test underscores ongoing complexities in detecting AI-generated media as generative models continue to advance. The inability of widely used chatbots to identify Sora-generated videos reflects broader challenges faced across the AI technology sector in content authentication. Efforts to develop specialized tools for media verification remain critical amid rising concerns about misinformation and synthetic content.

The market continues to see rapid innovations in AI-generated media tools alongside an increasing demand for verification capabilities. Competitors in the AI space are expanding the use of generative models for video and other multimedia, intensifying the need for more robust detection technologies.

As AI-generated content becomes more sophisticated, future developments will likely focus on enhancing the accuracy of detection models. Monitoring advancements in both generative AI and media verification tools will be key to understanding how the industry addresses these challenges moving forward.

Tests reveal that top AI chatbots including ChatGPT cannot reliably identify videos created by OpenAI’s Sora video generator.