On March 24th, OpenAI announced it would shut down Sora on April 26th. This announcement to the BBC came as a huge shock – why would a company with 900 million weekly users, worth almost $1 trillion, shut down a programme that was so rapidly gaining popularity and ability?

Most shockingly, this statement comes just a few months after Disney announced it was bringing 200 of its most beloved characters to Sora. This agreement was quoted at $1 billion. What on earth happened?

The Cost

Maintaining Sora is ridiculously expensive. For a customer to generate just one 10-second video, this costs OpenAI $1.30. With 1 million downloads in 5 days when Sora hit the market in September of 2025, there are upwards of 10 million videos created with the app daily. This is leading to OpenAI bleeding about $15 million DAILY. That comes out to around $5.2 billion yearly.

It’s a bit more than a daily coffee or weekly petrol top-up. In a tweet from Bill Peebles, a research scientist at OpenAI, ‘the economics are totally unsustainable’. As it stands, OpenAI isn’t expected to profit until 2029 – 14 YEARS after it was founded in December 2015.

The cost comes from the GPU, or Graphics Processing Unit. This is, in very basic terms, the computer power required to generate this content. With such a high demand, OpenAI wasn’t able to charge users sufficient amounts to cover costs, no matter how many pro users they had. Because of this, they were practically offering a free service that was as good as flushing money down the toilet.

OpenAI’s Upcoming IPO

An IPO is an Initial Public Offering. This is where a private company first sells shares to the public in order to raise capital. Until now, OpenAI has not been publicly traded, so this transition will allow it to fund further growth and scale even more. But how does this affect Sora?

Obviously, investors want to invest in a healthy business that they believe will reward them. In losing $15 million a day, OpenAI weren’t exactly painting this picture. In getting rid of Sora ahead of their IPO later this year/early 2027, they may appear healthier and therefore be in a better position when going public.

A Streamlined Approach

The launch of ChatGPT in November of 2022 skyrocketed OpenAI into the media and tech space more than ever before. In doing so, they were widely regarded as industry leaders, ahead of competitors such as Google with Gemini and Microsoft with Copilot.

In the years following, OpenAI have widened their scope with the introduction of Sora, DALL-E, ChatGPT Atlas and more. On the other hand, competitors such as Anthropic have opted to deepen their scope, rather than widen it.

In doing so, OpenAI has arguably spread itself too thin with multiple projects that are only partially developed, whereas competitors have focused on a small number of products that they are constantly developing. Anthropic’s Claude is a prime example of this, as they have introduced Claude Code and Claude Cowork.

Shutting down Sora is potentially the start of OpenAI taking a more streamlined approach to their operations, and focusing more on developing rather than ‘innovating’. This comes after lots of backlash from reputable talent agencies such as CAA saying that OpenAI are exploiters rather than innovators, and are a significant risk to their clients such as Doja Cat and Scarlett Johansson and their intellectual property.

Natural Decline

Sora’s monthly downloads have decreased from 3.2 million in November to 1.2 million in January. Outside of Sora, OpenAI are seeing a general decline in usage, decreasing from a 69% app-market share in January 2025 to 45% in 2026.

As of early 2026, there is a significant movement of people away from OpenAI, more specifically ChatGPT. The #QuitGPT movement has seen hundreds of thousands of users cancelling subscriptions and switching to other tools.

A few reasons for this movement could be:

• Military Contracts: OpenAI recently signed deals to bring its models into classified military operations
• Data Privacy: OpenAI trains its models using user data, which raises privacy concerns
• Decreasing Quality: there is a declining perceived value compared to competitors
• Better Alternatives: as time continues and competitors continue to improve their models, there is less reason to stick with OpenAI

Sam Altman

OpenAI’s CEO, Sam Altman, has undergone a lot of backlash from the media for quotes such as “It takes like 20 years of life and all of the food you eat during that time before you get smart,” (headlined as putting robots and humans under the same umbrella) and that the worst case scenario is “lights-out for all of us”.

Whilst these quotes could arguably have been taken out of context, being so opinionated as a founder has a large impact on the course of your business. Over time, headlines really do add up and create a reputation that is then hard to change.

In an age of cancel culture, the public’s perception of Altman is increasingly being reflected in the usage of OpenAI’s products. If this decline was to continue, this could have significant contributions to the potential end of OpenAI as a company.

Is this the end of OpenAI?

Some internet users argue that this is the beginning of the end for OpenAI, predicting that they will either run out of capital, or be bought by a competitor such as Meta or Microsoft before everything comes crashing down.

Currently, OpenAI is trying to be too many things at once. They’re operating everywhere, yet dominating nowhere. People argue that OpenAI must streamline their operations for any chance of survival, otherwise they will simply be left behind.