Sora, the AI video-generation tool developed by OpenAI, was shut down just six months after its public release, a move that quickly triggered speculation about the company’s motives. Some users questioned whether the app’s ability to accept facial uploads pointed to a broader data-collection strategy. However, reporting from the Wall Street Journal suggests a more practical explanation: the product was costly to maintain, saw declining engagement, and was diverting resources from areas with clearer returns.
Following its launch, Sora initially attracted attention and reached an estimated global user base of around one million. That early interest didn’t hold. Usage dropped to fewer than 500,000 users, indicating that the novelty of AI-generated video did not translate into sustained demand. At the same time, the infrastructure required to support the service remained expensive. Running Sora reportedly cost OpenAI մոտ $1 million per day, largely due to the computational intensity of generating video content. Each request relied on high-performance AI chips, which are both limited and expensive to operate at scale.
This imbalance between cost and engagement appears to have made Sora difficult to justify internally. While resources were being allocated to maintain and improve the platform, other parts of the AI market were evolving quickly. Competitors such as Anthropic were gaining traction, particularly among developers and enterprise users. Its product, Claude Code, reportedly attracted attention in segments that are typically more stable and revenue-generating, including software engineering teams.
Against this backdrop, OpenAI CEO Sam Altman opted to reallocate resources. Shutting down Sora freed up computational capacity that could be redirected toward more widely adopted services, including language models and developer-focused tools. The decision reflects a broader strategic shift toward prioritizing products that scale more efficiently and align with enterprise demand.
The shutdown also had immediate implications for partnerships. Disney had reportedly committed $1 billion to a collaboration tied to Sora, but learned of the product’s discontinuation less than an hour before the public announcement. With the platform no longer active, the agreement was effectively terminated.
Taken together, Sora’s short lifecycle highlights a recurring challenge in the AI sector: not all technically impressive products translate into sustainable businesses. High operating costs, combined with uncertain user retention, can quickly outweigh early excitement. In this case, OpenAI’s decision appears less about controversy and more about focusing limited resources on areas with clearer long-term value.
