OpenAI appears to be recalibrating its approach to copyright and intellectual property within its new AI video app, Sora, following a wave of scrutiny from Hollywood and rights holders.
The company, which launched the app this week to rapid popularity, is now signaling a pivot from its previously reported “opt-out” model — in which studios and agencies had to explicitly forbid the use of their intellectual property — toward an “opt-in” framework that would require permission before any copyrighted material is used in user-generated content.
A Swift Rise, and Swift Scrutiny
Despite being invite-only, Sora has surged to the top of Apple’s App Store rankings, emerging as the No. 1 free app within days of launch. The platform allows users to generate short AI videos from text or images, and its standout feature, “cameos,” lets users upload biometric data to place their digital likeness into generated scenes.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
That same feature has also fueled controversy. Social media is already filled with Sora-made clips featuring copyrighted characters like Pikachu and SpongeBob — in some cases, even portraying them alongside AI deepfakes of OpenAI CEO Sam Altman, with the characters ironically criticizing OpenAI’s approach to copyright.
The viral spread of such videos underscores the tension between innovation and infringement that generative AI firms are increasingly confronting.
Hollywood’s Pushback
Before the app’s launch, The Wall Street Journal reported that OpenAI had informed studios and agencies they needed to opt out if they didn’t want their content used in Sora-generated videos. That approach drew concern from rights holders who argued that the policy placed an undue burden on creators to protect their IP.
In a blog post on Friday, Altman appeared to confirm a change in direction, writing that OpenAI plans to give copyright holders “more granular control over generation of characters, similar to the opt-in model for likeness but with additional controls.”
The language shift — from opt-out to opt-in — marks a significant turn. It suggests OpenAI is now building mechanisms to block users from generating videos with copyrighted characters unless the rights holders explicitly allow it.
“Interactive Fan Fiction” or Copyright Minefield?
Altman maintained that many rights holders have expressed enthusiasm for what he called “a new kind of interactive fan fiction,” viewing Sora as a potential avenue for audience engagement. However, he added that studios and creators “want the ability to specify how their characters can be used — including not at all.”
Even with the new safeguards, Altman acknowledged that Sora will not be immune to misuse. “There will likely be some edge cases of generations that get through that shouldn’t,” he said.
Altman also hinted at OpenAI’s evolving business model for Sora. Initially, the company said it would only monetize the app by charging users during high-demand periods. But in his post, Altman indicated a broader plan, noting, “We are going to have to somehow make money for video generation.”
He added that OpenAI is exploring ways to share revenue with rightsholders whose characters or content contribute to user-generated videos, describing the goal as “a new kind of engagement that is even more valuable than the revenue share, but of course we … want both to be valuable.”
A Broader Industry Reckoning
OpenAI’s recalibration comes amid growing legal and ethical battles over how AI companies train and deploy models using copyrighted data. In recent months, media groups, including The New York Times, have sued OpenAI and Microsoft for alleged copyright infringement, while major music labels have pursued lawsuits against AI startups accused of cloning voices and melodies without authorization.
In the film industry, studios and talent agencies have expressed concerns that AI-generated content could blur ownership boundaries — especially if fans or creators use tools like Sora to produce derivative works that resemble copyrighted material.
The question of “fair use” remains largely unsettled, and regulators in the U.S. and Europe are still wrestling with how to classify generative AI’s relationship with existing copyright law.
For OpenAI, which is now balancing partnerships across entertainment, publishing, and software, Sora’s debut is both a breakthrough and a test case. The app’s early success demonstrates massive demand for AI-driven creative tools, but its viral misuse has already highlighted the fragile line between innovation and infringement.