AI Video Generators Like Sora Threaten Trust in Visual Media
OpenAI's Sora app exemplifies how AI-generated videos are complicating trust in visual evidence, prompting calls for critical assessment of video content.
- • OpenAI released Sora, an AI app generating realistic videos from text descriptions.
- • Sora enables both entertaining content and dangerous disinformation, such as fabricated crime footage.
- • Experts warn this technology challenges the reliability of video as objective evidence.
- • Society must develop critical skills to evaluate video content similarly to written media.
Key details
OpenAI's new app Sora, which enables users to generate highly realistic videos from typed descriptions, is reshaping perceptions of video as credible evidence. Released in 2025, Sora allows users to create anything from humorous clips to disturbing fabricated footage, including fake crime scenes like "police bodycam footage of a dog being arrested."
While Sora's creative potential is lauded, its misuse in spreading disinformation—such as fabricated security footage of non-existent crimes—is fueling concern about the erosion of media trust. Experts warn that as AI-generated videos become easier and more convincing, the longstanding assumption that video is an objective record of reality must be reconsidered.
Ren Ng, a computer science professor at UC Berkeley, emphasizes the need for society to develop critical viewing skills similar to how readers scrutinize written material. This shift could profoundly alter how visual media is consumed and trusted, challenging journalists, legal professionals, and the public to differentiate genuine footage from AI fabrications.
The emergence of AI video generators from tech giants like Meta, Google, and OpenAI signals a significant transformation in media literacy. As these tools democratize video production, the line between authentic and artificial imagery blurs, demanding more sophisticated verification techniques.
This concern parallels wider discussions on AI's societal impacts, including shifts in human cognition and agency explored at recent academic panels. However, the immediate challenge presented by video generators like Sora lies in balancing technological innovation with safeguarding truth and public trust in an era where seeing is no longer believing.