Over the past decade, the internet has witnessed an explosion of video-sharing platforms and formats. It began with TikTok, which launched in China in 2016, made its way to India, and became a crowd favourite before being banned over national security and privacy concerns. Thereafter, Instagram Reels and YouTube Shorts emerged. And now, OpenAI has rolled out Sora.
Sora is OpenAI’s text-to-video generation tool that transforms written prompts, images, or existing video clips into high-quality, realistic videos. OpenAI launched the app last week—Sora 2.0’s enhanced realism, speed, and creative control makes it one of the most advanced AI video platforms available today. The underlying technology debuted last year, but its latest version, which is faster and more powerful, can incorporate your likeness if you upload images of your face.
Sora 2.0 allows users to generate videos from text prompts, images, or existing clips with stunning realism and speed. It comes packed with tools to remix, re-cut, loop, and blend videos, giving creators control over every frame. Its built-in storyboard editor lets users assign different styles and camera angles to each scene, while curated presets offer everything from film noir to surreal animation. It also comes with built-in audio generation, flexible output formats, and a TikTok-style vertical feed.
Rohan Sahai, OpenAI’s product lead for Sora, said in an interview: “We felt the best way to bring this technology to the masses is through something that is somewhat social. When you have such drastically shifting technology—a new form factor, even—our goal and philosophy as a company is to get it out there.”
Also Read: ChatGPT’s app integrations signal shift toward AI-powered super-app
The app is available on an invite-only basis only on iOS’s devices, which means users need a code to download it. Since its launch on September 30, the app has already climbed to the top of Apple’s App Store, as per reports.
Reports suggest that users are having a great time exploring the platform, with many videos gaining traction simply for their sheer creativity and fun. According to CNBC, which got early access to Sora, the platform is already brimming with clips featuring characters from popular shows like SpongeBob SquarePants, Rick and Morty, and South Park, as well as movies like Despicable Me. “People are eager to engage with their family and friends through their own imaginations, as well as stories, characters, and worlds they love, and we see new opportunities for creators to deepen their connection with the fans,” said Varun Shetty, OpenAI’s head of media partnerships, in a statement to CNBC.
It’s the use of copyrighted characters that has stirred up serious debate around the app. Experts warn that many of the videos people are creating using well-known cartoon characters are likely to cross into copyright infringement territory. Mark Lemley, professor at Stanford Law School, said in an interview that OpenAI is “opening itself up to quite a lot of copyright lawsuits.” Characters are protected intellectual property, meaning third parties can’t use them without permission—and Sora, with its open-ended creative tools, could easily become a hotspot for legal disputes.
This isn’t the first time AI platforms have faced scrutiny over copyright. Disney and Universal have already sued image generator Midjourney for allegedly misusing characters from their films.
Sora does have some restrictions in place. Rights holders must opt out on a case-by-case basis using a copyright disputes form, and public figures can choose to grant permission for their likenesses to be used.
According to a report by The New York Times, Hollywood has been on edge since Sora’s release, concerned about how easily users can replicate likenesses without compensation. A day after launch, talent agency WME sent a memo to its agents stating: “There is a strong need for real protections for artists and creatives as they encounter A.I. models using their intellectual property, as well as their name, image and likeness.” The agency also informed OpenAI that all its clients were opting out of having their likenesses or IP used in Sora-generated content.
In response to all these claims, OpenAI said it honours takedown requests submitted through its “Copyright Disputes” form, which allows rights holders to flag specific content. Users can also report videos for copyright or trademark infringement directly within the app. The company clarified that these actions are handled on a character-by-character basis, rather than through blanket bans.
There is also another important aspect to consider. According to Meghna Bal, director of the Esya Centre, a technology policy think tank based in New Delhi, an animation style itself is not protected by copyright. She believes that protecting a style would restrict other artists’ ability to create and would inhibit creative expression. “Extending copyright protection beyond a specific work to include its stylistic elements would threaten the creative ecosystem by limiting future artists’ capacity to draw inspiration from existing art,” says Bal.
She further explains that the boundary between “style” and “expression” is not always clear cut. Artists often cultivate distinctive styles to set their work apart, and some US courts have acknowledged that style can play a role in copyright analysis. Under Indian copyright law, she adds, AI-generated outputs are only infringing if they meet the threshold of substantial similarity laid down by the Supreme Court in RG Anand v Deluxe Films. In this case, the Court held that for a claim of copyright infringement to be actionable, the copy must be a “substantial and material” one. It added that this threshold is met if a viewer or a reader gets an “unmistakable impression” that the subsequent work is a copy of the original work. If there are “material and broad dissimilarities” between the two works that negate the intention to copy the original, no infringement is made out.
Broadly, according to Bal, AI-generated outputs could either be (i) nearly identical to the training data; (ii) similar to it in some ways but different in others; or (iii) very different from the training data. The outcome of the substantial similarity assessment is clear in the first and third scenarios. However, it is the second scenario that is likely to require closer assessment and pose challenges for policymakers and courts alike.
“From what I have seen, Sora has introduced controls that allow users to control what their cameos say and do. In terms of consent, and withdrawal of consent, the DPDPA has been enacted but is not yet in force. But if a user wants to misuse someone’s likeness and distribute it, the liability should not automatically fall on the platform. In such cases, the users must be held liable,” says Bal. She explains that, in such cases, criminal enforcement provisions would come into play. For instance, Sections 66C, 66D, and 66E of the IT Act, respectively, make identity theft, cheating by personation, and violation of privacy punishable offences. “We should resist the temptation to make situations like these subject to compliance, as it takes away incentives from platforms to innovate around how to better protect users,” concludes Bal.
First Published: Oct 08, 2025, 18:40
Subscribe Now