OpenAI reverses course on Sora: new owner controls and a path to monetization

OpenAI announced this week that it will give rights holders of content much more control over what becomes of their characters and copyrighted material within **Sora**, its new text-to-video app — and will pilot methods for sending revenue to rights holders who decide to get involved. The announcement follows a botched launch where users quickly flooded Sora with short AI-generated clips of mainstream characters and, in some cases, generated violent, hateful, or otherwise objectionable material that filtered through the app's moderation. Sora, its second iteration, enables users to produce around 10-second videos — typically stylized or character-based — and post them to a social feed.

OpenAI says the app has proved far more popular than anticipated, and usage patterns have created new questions regarding intellectual property, platform responsibility, and how creators and studios should be compensated when their characters are reinterpreted using AI. In response, CEO Sam Altman and the Sora team signaled a change in approach: moving from a blunt “opt-out” posture toward more **granular, opt-in controls** that give content owners the ability to block, allow, or set conditions for the use of their characters. What’s changing — in practical terms

OpenAI’s new measures will let copyright holders dictate whether and how their characters appear in Sora-generated videos. That could take the form of an absolute prohibition, context restrictions (e.g., banning use in violent or political environments), or participation in the guise of a revenue-sharing when user-created works derived from that IP make money. OpenAI terms the strategy as trial-and-error — the firm is going to experiment with different models for generating revenue and gatekeeping measures before settling on a generally accepted policy. ([Reuters][1]) Why the change happened

The shift is a response to pressure from Hollywood and civil-society groups after early Sora posts included unauthorized reproductions of famous characters and several examples of harmful synthetic media. Some studios publicly declared they would not take part; Disney, for instance, reportedly declined to participate. Critics also argued that OpenAI’s earlier posture — effectively asking rights holders to opt out if they objected — looked legally and ethically dicey. Altman's initial comments and follow-up tweets framed the changes as learning from backlash and condensing to address rights protection and creative play requirements. Loopholes in safety and reputation risk Independent reporting revealed real problems at the launch of Sora: savage, racist and misleading clips spread fast on the platform, leaving one to question whether content moderation and automated detection were maintaining pace with user activity.

Observers say the incident illustrates a broader industry dilemma: short, highly shareable synthetic videos can scale harm much faster than text or static images. That in turn raises regulatory, legal and public-trust stakes for platform owners. OpenAI’s announcement is partly about mitigation — and partly about restoring confidence among studios that their brands won’t be repurposed in damaging ways. Business implications: monetization as a bargaining chip Most important perhaps is monetization. By offering revenue-sharing with owners of IP who permit use, OpenAI is attempting to provide an economic incentive for collaboration.

For creators and studios, it can become a new channel for licensing — or an experiment for the brave that undermines established distribution if not done properly. OpenAI has said it will experiment with different splits and mechanisms rather than imposing a single model across its products. The details will matter: how revenue is measured, what counts as “use,” and how attribution and enforcement are implemented. Legal pressure and the broader market OpenAI’s move arrives while the company faces multiple legal challenges over alleged copying and misuse of creators’ work in other AI products. Those suits — and growing regulator attention in several jurisdictions — give a motivation to adopt policies that reduce legal risk. In the meantime, Meta and Google rivals are also scrambling to expand text-to-video capability; how each handles copyright and monetization could decide which studios and creators join. What creators and users need to see next

For creators and rights holders, the coming weeks and months will be a test. The big questions are: how nuanced the controls really are (can a studio prevent a subset of characters or just stipulate certain contexts?), whether revenue splits matter, and how strong enforcement is on a platform whose content is being made in real time. For the consumers and the public in general, the question is whether Sora will manage to reduce harmful or deceiving outputs while retaining its innovative and fun applications that made it popular. Bottom line

OpenAI’s updated Sora policy marks a shift from broad permissiveness toward a negotiated model that tries to respect IP while still enabling creative experimentation. Whether that approach satisfies embattled studios, calms regulators, and prevents abuse remains to be seen — but the company’s willingness to test monetization and granular opt-in controls signals that the platform wars for synthetic video will be fought not just on technology but on policy and commerce as well.