OpenAI Is Discussing Bringing Sora Back to Life Six Months After Shutting It Down

About six months ago, OpenAI shut down Sora, and the story seemed to be over. On October 13, 2026, it resurfaced: according to an exclusive from The Information, people inside the company are discussing a plan that could "resurrect" the video model. The article contains no confirmation from OpenAI itself. It describes a discussion, not a decision.

The key detail is who is in line to take the project. The Chinese outlet Jiqizhixin, summarizing the report, writes that the team has a "strong background." It includes a well-known Hollywood producer and people from DreamWorks. The available excerpt does not give a full list of names, and naming anyone based on a guess would be wrong.

What we know so far

There are few facts, so it is best to lay them out clearly:

• Sora was shut down about six months ago, according to the wording of Jiqizhixin's headline.
• The source of the news is an exclusive from The Information; the summary is dated October 13, 2026.
• OpenAI is discussing a revival plan, and there is no final decision.
• The potential recipients of the project include a Hollywood producer and DreamWorks alumni.
• The format of the deal is unknown: it could be a license, a spin-off, an asset transfer or a joint venture.
• The timeline, the amount of funding and the fate of the technology inside OpenAI have not been disclosed.

Why Hollywood

From the start, Sora was a product at the intersection of AI and film. The first version, shown in 2024, produced clips that people compared to commercials and short films. Sora 2 launched in late September 2025 with a standalone app, and it immediately sparked disputes over copyright and the use of people's likenesses. The film industry was both interested in the model and wary of it.

That makes the lineup of candidates look logical. Producers and studio specialists know how to work with rights holders, how to build a production pipeline and where the legal boundaries lie. For a lab whose main focus has shifted to language models, agents and infrastructure, this is a convenient way to avoid abandoning the area for good. Video generation remains expensive in compute, and commercializing it takes partners who understand the customer better than researchers do.

The sources do not explain why now. One can build various theories, but for now they are only theories. It also cannot be ruled out that the idea will not reach a deal: a discussion remains a discussion.

What comes next

There are more questions than answers. Will the model stay under the OpenAI brand? Will the new team get access to the weights and training data? Who will handle the copyright issues that Sora 2 raised? The answers will determine whether this is the return of a product or the handover of a technology to outside hands.

What to do now

Teams that built workflows on Sora should treat the news with caution. Do not build the service's return into your plans until OpenAI makes an official statement. Save your exported materials and prompts, and keep your working pipelines on alternative video models. Studios and agencies should follow The Information's reporting and OpenAI's own announcements: if the deal goes through, access and licensing terms will be the first things to change. Investors and analysts should note the precedent itself: a shuttered AI product from a major lab can get a second life outside that lab.