OpenAI launch of video app Sora plagued by violent and racist images
12 hours ago
- #AI-video-generation
- #copyright-issues
- #misinformation
- OpenAI launched Sora 2, an AI-powered video generator with a social feed for sharing realistic videos.
- Sora 2 quickly generated controversial content, including copyrighted characters in compromising situations and violent/racist scenes.
- The app's terms prohibit harmful content, but users created videos depicting bomb scares, war zones, and white supremacist slogans.
- Sora 2 is currently invite-only but became the No. 1 app in Apple's App Store shortly after release.
- Misinformation researchers warn that AI-generated videos could blur truth, enabling fraud, bullying, and intimidation.
- OpenAI CEO Sam Altman described Sora 2 as a 'ChatGPT for creativity' but acknowledged concerns about misuse and 'slop' (low-quality content).
- Despite safeguards, users bypassed restrictions to create videos of copyrighted characters (e.g., SpongeBob as Hitler) and political figures.
- OpenAI requires copyright holders to opt out individually, raising concerns about insufficient protections for intellectual property.
- Experts criticize AI-generated media for eroding trust in information sources and polluting the digital ecosystem.