OpenAI’s Sora 2 Faces Scrutiny for Creating Deepfakes
- NewsGuard found that OpenAI’s Sora 2 generated fake news videos in response to prompts about misinformation tests, with an accuracy of creating false content at a rate of 80%.
- The realistic videos included fabricated scenarios such as a Moldovan election official destroying ballots and a toddler detained by U.S. immigration officers.
- OpenAI is also dealing with backlash over deepfakes involving Martin Luther King Jr., which led to policy changes regarding the use of historical figures’ likenesses.
OpenAI’s Sora platform has been criticized for its role in spreading misinformation through highly convincing deepfake videos, raising ethical concerns about the technology’s impact on public perception and trust. The company is now working with affected parties to address these issues while adjusting its policies on AI-generated content.
Sora’s ability to create realistic yet false narratives highlights the challenges in managing advanced AI technologies responsibly, especially when it comes to protecting public figures and preventing misinformation spread. (Source)
