OpenAI is limiting deepfakes on Sora after pressure from Bryan Cranston and Hollywood groups

In a significant move, OpenAI has announced it will collaborate with actor Bryan Cranston and various Hollywood organizations to impose restrictions on deepfake content generated through its Sora 2 video application. This decision comes after Cranston raised alarms regarding unauthorized uses of his likeness and voice in the video generator, which was launched this fall on an invite-only basis. Cranston expressed his worries not only for himself but also for fellow performers whose identities could be exploited in similar ways. His concerns prompted action from SAG-AFTRA, among other organizations, leading to a joint statement that highlights the collaborative effort to protect artists. According to OpenAI, the newly implemented policy mandates that individuals must explicitly opt-in to allow their names and likenesses to be used by Sora. The company acknowledged past instances of unintentional misuse and stated that it has enhanced controls to prevent unauthorized replication of individuals' voices and images. The statement emphasized that all creators should have the authority to dictate how and when their likenesses are represented in AI-generated content. OpenAI also assured that it is dedicated to addressing any grievances promptly. This policy adjustment aligns with the proposed NO FAKES Act, which aims to safeguard individuals' rights concerning their voice and likeness against AI misuse. Since the rollout of Sora 2, OpenAI has already begun implementing changes to bolster protections for intellectual property holders in response to public outcry regarding copyright issues. Notably, the company recently announced a pause on AI-generated videos featuring civil rights leader Martin Luther King Jr. after his estate raised concerns over potential misrepresentation.