TLDR: OpenAI’s video-generating app, Sora, has rapidly achieved over 1 million downloads in under five days, outperforming ChatGPT’s initial growth. Despite its swift adoption, the invite-only iOS app, currently limited to the U.S. and Canada, is experiencing mixed user feedback, reflected in a 2.8-star App Store rating, stemming from issues like content moderation, AI prompt generation frustrations, and ethical concerns over deepfake capabilities.
OpenAI’s groundbreaking video-generating application, Sora, has achieved a remarkable milestone, surpassing one million downloads in less than five days since its launch on September 30, 2025. This rapid adoption rate outpaces even OpenAI’s highly successful chatbot, ChatGPT, during its initial rollout. The achievement is particularly notable given that Sora is currently an invite-only application, exclusively available to iOS users in the United States and Canada.
Despite its impressive download figures, Sora is navigating a landscape of mixed user reception and growing controversy. The app currently holds a 2.8-star rating on the Apple App Store, a reflection of user frustrations. Critics point to several factors contributing to this lukewarm feedback.
One significant area of concern revolves around the inherent challenges of AI prompt generation, with some users experiencing ‘botched videos’ despite highly specific instructions. Additionally, the app has faced criticism for its stringent content censorship. Initially, Sora sparked debate over its controversial opt-out model for copyrighted material, which required creators to explicitly disallow their work from being used in user-generated Sora videos. OpenAI later reversed this to an opt-in model.
Ethical considerations surrounding the creation of deepfake videos have also come to the forefront. Reports indicate the circulation of AI-generated videos featuring deceased public figures, including Michael Jackson and Tupac Shakur, raising serious questions about consent and the digital use of likenesses. Zelda Williams, daughter of the late Robin Williams, has publicly urged users to refrain from sharing AI-generated videos of her father, highlighting the broader ethical dilemmas posed by Sora’s capabilities. This issue is part of a larger trend, with several companies developing generative AI systems currently facing lawsuits over the use of copyrighted materials in their training datasets.
Also Read:
- Japan Urges OpenAI to Address Anime and Manga Copyright Concerns with Sora 2
- Veo 3.1 Elevates Google’s AI Video Capabilities, Posing Direct Challenge to Sora 2
Sora allows users to convert text prompts into realistic ten-second videos and supports audio generation, adding another layer of realism to its creations. A unique ‘cameos’ feature enables users to insert their own likeness, or that of a friend, into an AI-generated video. OpenAI’s Sora lead, Bill Peebles, has acknowledged the surging growth and user feedback, stating that the team is ‘working hard to keep up’ and that new features and fixes, including addressing issues of over-moderation, are forthcoming.
