Sora 2: immense popularity & controversies – AI News – #2 October 2025

The latest tool from OpenAI, Sora 2, has shaken the tech world. The AI video generation app not only climbed to the top of download charts but simultaneously sparked a wave of discussions about ethics, copyrights, and the future of the internet. We analyze this dual picture: on one hand a spectacular success, on the other – chaos and serious accusations.

Sora 2 is the latest generation of OpenAI’s artificial intelligence model designed to generate videos based on text descriptions. In practice, this means that a user can enter a command (a so-called “prompt”), and the AI will create a video clip that tries to faithfully depict the described scene.

Compared to its previous version, Sora 2 offers a significantly higher level of realism, better control over generated content, and longer, more coherent shots. At the same time as the model’s launch, OpenAI released an iOS app that functions as a social platform. Users can not only create videos but also share them, including generating deepfakes featuring themselves or their friends, which from the start was one of the key and most controversial elements of the service.

Even with limited access (currently by invitation only in the USA and Canada), the launch of the Sora 2 app can be described as a market phenomenon. The level of interest exceeded many analysts’ expectations and showed how huge the consumer appetite is for advanced AI tools put into their hands.

The statistics speak for themselves. According to data from Appfigures:

Sora 2’s debut was stronger than the launches of AI apps like Claude (Anthropic) or Copilot (Microsoft). Considering the limited access, these numbers suggest that if the app were open to everyone, its initial success would have been even more spectacular.

Sora 2’s popularity proves one thing: users no longer just want to read about AI capabilities. They want to use them in a simple, intuitive, and—most importantly—creative way. OpenAI has packaged powerful technology in a format resembling a social network, which naturally encourages experimentation and sharing of results. This is a recipe for viral success.

Unfortunately, behind the impressive stats lies a story full of stumbles that casts a shadow over the entire venture. The launch of Sora 2 quickly turned into a PR crisis for OpenAI, forcing the company to put out fires on several fronts simultaneously.

OpenAI’s initial decision to apply an “opt-out” model for copyright issues proved catastrophic. This meant the model could by default use images and intellectual property unless the rights owner explicitly objected. The result? The internet was flooded with problematic content such as an image of Nazi SpongeBob, ads for fictional “Epstein Island” toys, and… a likeness of Stephen Hawking performing tricks like Tony Hawk:

After strong backlash from the media and organizations including the Motion Picture Association, OpenAI withdrew this policy and switched to a much safer “opt-in” model, requiring explicit consent to use images. The company now tries to encourage creators to participate, promising them greater control over how their “cameos” will be used in generated videos.

Digital security experts point to an even deeper and more disturbing problem. Sora 2, in its accessible form, effectively changes the perception of deepfake technology—from a dangerous disinformation tool to a lighthearted, entertaining toy. As Daisy Soderberg-Rivkin, former TikTok Trust & Safety manager, put it: “It’s as if deepfakes have gained a publicist and a distribution deal.”

This sudden normalization and mass dissemination of ultra-realistic fake video content may lead to a sharp decline in trust towards online content. In a polarized world, the ability to “prove” any claim using a convincingly realistic video is a scenario for which society is not prepared.

The case of Sora 2 is a perfect illustration of the dilemmas facing the entire AI industry. On one hand, there is breakthrough technology with enormous creative and commercial potential. On the other—huge risks when such a powerful tool is released to the market without proper preparation and consideration of consequences.

OpenAI is now racing against time—trying to patch security gaps while users are already finding ways to circumvent them. The pressing question remains: will the company manage to control the chaos and turn a controversial launch into a long-term success? One thing is certain: the boundary between truth and fiction on the internet has just been shifted, and we will all have to learn to operate in this new reality. If you want to stay up to date with changes in the AI world, make sure to subscribe to the Delante newsletter!

Sources of information about Sora 2: