OpenAI is facing mounting legal pressure and plans to give copyright owners more granular control over how their works appear in Sora 2-generated videos.
OpenAI has announced it will give copyright holders more control over how their intellectual property is used in videos produced by Sora 2. The shift comes amid criticism over Sora’s ability to generate scenes featuring popular characters and media, sometimes without permission.
At launch, Sora allowed generation under a default policy that required rights holders to opt out if they did not want their content used. That approach drew immediate backlash from studios and creators complaining about unauthorised use of copyrighted characters.
OpenAI now says it will introduce ‘more granular control’ for content owners, letting them set parameters for how their work can appear, or choose complete exclusion. The company has also hinted at monetisation features, such as revenue sharing for approved usage of copyrighted content.
CEO Sam Altman acknowledged that feedback from studios, artists and other stakeholders influenced the change. He emphasised that the new content policy would treat fictional characters more cautiously and make character generation opt-in rather than default.
Still unresolved is how precisely the system will work, especially around the enforcement, blocking, or filtering of unauthorised uses. OpenAI has repeatedly framed the updates as evolutionary, acknowledging that design and policy missteps may occur.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!