OpenAI's Shutdown of Sora Could Be a Warning Signal for the AI Video Market

OpenAI has wound down Sora, and this is more than a corporate reshuffle. The closure of one of the most anticipated AI video products in recent years forces an uncomfortable question on the entire industry: have we overestimated the potential of generative video?

Sora debuted in February 2024 as a demonstration of a new level of capability. Clips featuring realistic people, physically plausible scenes and smooth transitions stunned the public. OpenAI positioned the product as a breakthrough comparable to the arrival of ChatGPT in text generation. By December 2024, the model was released publicly, first to ChatGPT Plus and Pro subscribers.

But behind the loud announcements lay problems that were never fully solved. Video generation demands enormous computing resources, far more than text or images. Wait times remained significant. Monetization never reached break-even. Meanwhile, competitors such as Runway, Kling, Pika and Hailuo did not stand still, offering comparable quality at a lower price.

Now that Sora has been taken out of service, industry observers have split into two camps.

The first sees normal corporate strategy: OpenAI is likely reallocating resources toward higher priorities such as GPT-5, agent systems and enterprise solutions. Closing a product does not mean abandoning video altogether; the technology may be integrated into other services or relaunched in a different form.

The second camp is less optimistic. In their view, Sora's fate is a symptom of a systemic problem. AI video turned out to be more expensive than expected, in both production and infrastructure. Real demand from paying users proved lower than interest metrics and the viral views of demo clips suggested. Brand hype did not convert into sustainable revenue.

Historically, this is not the first time a high-profile AI product has failed to meet commercial hopes. Google has relaunched and renamed its AI assistants several times. Meta has shut down ambitious metaverse projects. The gap between impressive demos and real product viability is a long-standing ailment of the tech industry.

For the AI video market as a whole, Sora's closure is a maturity test. Venture money flowed actively into startups in the segment throughout 2024–2025. Runway raised funding at a multibillion-dollar valuation. Chinese players aggressively entered the global market. Now investors and founders will have to answer a question that was convenient to postpone until now: who is willing to pay, and for what, in this segment?

The consumer market has not yet formed a lasting habit around AI video the way it did with text assistants. The professional market (film, advertising, media) is interested in the tools, but slowly and cautiously. Regulatory pressure, copyright questions and ethics add further friction.

What this means in practice: a wave of consolidation in AI video is inevitable. Small players without a differentiated monetization model will come under pressure. The survivors will be those who have found a specific professional niche, such as ad production, education or games, and built a sustainable business around it.

Sora's closure does not kill the industry, but it will certainly force it to grow up.

*Meta is recognized as an extremist organization and banned in Russia.