A recent investigation by Newsguard highlights a significant shortcoming in the capabilities of leading chatbots, including ChatGPT, which reportedly fails to recognize 92% of fake videos generated by OpenAI's Sora tool. This finding raises important questions about the reliability of AI systems in discerning authentic content from manipulated media. In this context, it is crucial to consider the broader implications of such failures, particularly as misinformation continues to proliferate online. While the technology holds promise for various applications, the inability to accurately identify deceptive content poses risks for users and stakeholders alike. A final assessment of the situation would be premature, as ongoing advancements in AI may eventually address these challenges. However, the current limitations underscore the need for vigilance and further research in the field.