OpenAI's Sora
OpenAI shuts down Sora app following fading interest and user backlash AFP News

OpenAI has announced it is shutting down its Sora social video app, just months after its public rollout, following a decline in user engagement and ongoing concerns about the platform's content. The company confirmed the decision on 24 March but did not provide a detailed explanation or timeline for the app's full closure. The app had also faced criticism from some users, who described its highly realistic AI-generated videos as 'creepy' or unsettling.

Sora was designed as an AI-first social platform built around short-form video, allowing users to generate clips using OpenAI's video model. The app initially launched with strong interest through an invite-only system, but struggled to maintain momentum after its wider release. The decision to discontinue the app marks a shift away from consumer-facing video products, even as the underlying Sora model continues to be developed within OpenAI's broader tools.

The move comes as the company increasingly focuses on integrating its AI systems into existing platforms. While OpenAI has not framed the shutdown as a strategic pivot, the closure follows months of mixed reception and slowing growth.

How the Sora App Worked

Sora functioned similarly to a short-form video platform, combining AI-generated visuals with a scrolling feed format familiar to users of apps such as TikTok. One of its main features allowed users to create realistic digital likenesses of themselves, which could then be used to generate videos.

The app attracted early attention for the quality of its output, but also raised concerns about how easily the technology could be used to create misleading or unauthorised content. Although safeguards were introduced, users were able to generate videos resembling public figures and fictional characters, prompting criticism from rights holders and public figures.

Reports also noted that some content circulating on the platform included deepfakes of well-known individuals, as well as the use of copyrighted material without permission. These issues contributed to questions about how such tools could be moderated at scale.

Declining Downloads and High Costs

Despite initial interest, Sora did not sustain long-term user growth. Data from AppFigures shows the app reached more than three million downloads at its peak in November 2025, before declining to around 1.1 million downloads by February 2026.

While these figures indicate continued usage, they fall short when compared with the scale of OpenAI's core products. ChatGPT, for example, has hundreds of millions of weekly users, highlighting the gap between experimental applications and established platforms.

The app also generated limited revenue through in-app purchases tied to video generation credits. At the same time, the computational cost of producing AI-generated video remained high, making it difficult for the platform to justify continued investment without sustained growth.

Partnership Plans and Industry Concerns

The shutdown also affects a previously announced collaboration involving entertainment content. According to reporting cited by TechCrunch, a proposed agreement that would have allowed the use of major film and television characters within the app will not proceed following the closure.

Some developers and creators viewed Sora as an early demonstration of how generative video could be integrated into social platforms. Others raised concerns about the ethical and legal risks associated with widespread access to realistic video generation tools.

Sora Model Continues Beyond the App

Although the Sora app is being discontinued, the underlying video generation model remains part of OpenAI's wider ecosystem. The company continues to develop the technology for use within its existing products, where access can be more tightly controlled.

The closure reflects the difficulty of translating early interest in generative AI into sustained engagement on standalone platforms. It also underscores the balance companies must strike between innovation, user demand and the risks associated with increasingly realistic synthetic media.