Overview
- OpenAI issued a joint statement with SAG-AFTRA, CAA, UTA and the Association of Talent Agents pledging stronger opt-in controls for performers’ voice and likeness after unauthorized Bryan Cranston generations during Sora 2’s invite-only launch.
- The company paused depictions of Martin Luther King Jr. at the request of his estate and says it is reinforcing safeguards for historical figures.
- Reality Defender reports it bypassed Sora’s anti-impersonation checks within 24 hours using publicly available footage, a claim OpenAI answered by saying it is continually hardening the system.
- Researchers and watchdogs highlight harmful content on the Sora feed, including antisemitic videos that are being widely remixed, while third-party tools can strip Sora’s visible watermark despite embedded provenance signals.
- Sora’s app surpassed about 1 million downloads in under five days, and creating videos requires a paid ChatGPT Plus account even as OpenAI promotes layered safety systems and support for the No FAKES Act.