Failing Sora App: OpenAI Plans to Fold AI Video Maker Into ChatGPT

A 45 percent crash in downloads during January 2026 has reportedly pushed OpenAI to abandon Sora as a standalone video generator, with plans to fold it directly into ChatGPT instead. The move would mirror how the company previously integrated image generation into its flagship chatbot, giving ChatGPT's hundreds of millions of weekly users access to AI video creation without switching apps.

Even a high-profile Disney collaboration couldn't prevent Sora's steep decline, highlighting the challenge of turning AI video generation from a social media novelty into a daily habit. Users downloaded out of curiosity but had little reason to return once the initial excitement wore off.

The integration promises broader reach but amplifies existing headaches. Video generation costs far more computationally than text or images, meaning OpenAI would face a massive infrastructure bill serving ChatGPT's vast user base. Safety risks would also scale dramatically, particularly around deepfakes and unauthorized content that plagued Sora's early days.

OpenAI hasn't officially confirmed the plan or revealed how Sora access would work within ChatGPT's subscription tiers. The company faces unresolved questions about whether tighter content policies would accompany the integration, and whether giving Sora the audience it failed to build independently will solve underlying retention problems or simply relocate them to a more expensive, higher-stakes platform.