Sora App: Deepfake Yourself with OpenAI’s AI
- On Tuesday, October 1, 2025, OpenAI released Sora 2, an AI video request for iOS devices.
- Sora 2 is powered by OpenAI's latest video generation model, also named Sora 2.
- The process of video creation within Sora 2 is designed to be user-friendly.
“`html
OpenAI Launches Sora 2: A TikTok-Like AI video App
Table of Contents
Published October 1, 2025, at 20:30:39 PDT
What is Sora 2?
On Tuesday, October 1, 2025, OpenAI released Sora 2, an AI video request for iOS devices. The platform distinguishes itself by integrating AI-generated sounds directly into the videos, a first for OpenAI product releases. Currently, access is limited to users with invite codes, creating an initial exclusivity around the application. Sora 2 operates on a “For You” page format, similar to TikTok, showcasing a stream of user-created clips.
Sora 2 is powered by OpenAI’s latest video generation model, also named Sora 2. The model allows users to create videos from text prompts, and now, to incorporate the likenesses of individuals through a “cameo” feature. Users simply tap on faces within the app to add them as characters, then input a prompt to generate a scene.
Creating Videos and the “Cameo” Feature
The process of video creation within Sora 2 is designed to be user-friendly. Users can generate personalized deepfakes by utilizing the “cameo” feature. This allows for the inclusion of specific individuals in generated videos. For example, a user could input the prompt “fight in the office over a WIRED story” after adding faces to the generation page, resulting in a video depicting those individuals engaged in the described scenario.
During testing by WIRED,videos generated by Sora 2 exhibited some imperfections and errors. Though, the platform demonstrates a significant leap in creating seamless and convincingly realistic deepfakes. This raises crucial considerations regarding the potential for misuse and the need for responsible AI growth.
Security Concerns and Ethical Implications
The ease with which Sora 2 can generate realistic deepfakes has already sparked security concerns. WIRED reported an incident where a character within a generated video attempted to acquire a GPU,and when caught,a voice resembling OpenAI CEO Sam altman pleaded with a security guard to allow the character to retain the GPU for AI tool development. This anecdote highlights the potential for AI-generated content to be used in deceptive or manipulative ways.
The ability to create convincing deepfakes raises ethical questions about consent, misinformation, and the potential for reputational damage. The accessibility of Sora 2, even in its current limited release, underscores the urgency of developing robust safeguards and ethical guidelines for AI-generated content.
Technical Specifications and Future Development
Sora 2 currently operates exclusively on iOS devices. The platform’s reliance on an invite code system suggests a phased rollout,allowing OpenAI to gather user feedback and refine the model before a wider public release. The integration of AI-generated sounds represents a significant advancement in video generation technology, adding another layer of realism to the created content.
OpenAI has not yet announced specific plans for expanding Sora 2 to other platforms or adding new features. However, given the rapid pace of development