OpenAI is shutting down Sora, the viral video-generation application, just 6 months after its launch. It’s happening despite the app hitting 1 million downloads in just 5 days. The company has reported a simultaneous $11.5 billion loss for the last quarter in 2025 and is now making damage control. All these moves come as OpenAI is now choosing to retreat from the costly projects, while it prepares for a potential IPO, facing some mounting legal challenges from the co-founder Elon Musk and Microsoft, its largest shareholder.
Mounting losses make OpenAI retreat from its ambitious project Sora
The decision to shutter Sora marks OpenAI’s sharp reversal from December, especially when Disney agreed to invest $1 billion and license characters for the platform. However, this deal did not close. One of the spokespersons said that the company respects the decision of OpenAI for exiting “video generation business and to shift its priorities elsewhere.”
Closure of Sora follows a retrenchment pattern. Recently, OpenAI abandoned its Instant Checkout shopping feature. The company took a backstep from its announced plans for consolidating browser, Codex coding and ChatGPT application to one single desktop application.
There are some severe financial pressures mounting on OpenAI. The company pulled in approx. $13 billion in revenue in 2025. The company is also projecting losses of $14 billion in 2026. To run a video-generation service without any revenue stream just proved unsustainable, as the AI tech demands way more computing power as well as electricity than any traditional internet service.
The market share of ChatGPT collapsed from 69% in January 2025 to 45% by January 2026. In the meantime, Anthropic captured 8 of the top 10 Fortune 100 companies, as the customers of it are growing 10x YoY with a run rate of $20 billion in revenue.
Broken commitments and missteps drive OpenAI into a downward spiral

OpenAI has entered 2026, confronting the broken promises pattern, which transformed its allies into adversaries. $13 billion was invested by Microsoft under a cloud exclusivity agreement. However, OpenAI, with Amazon Web Services, signed $50 billion deal on February 27, 2026. It granted AWS exclusive hosting rights for OpenAI’s Frontier AI agent platform. Now, Microsoft warns, it is prepared to take legal action against OpenAI. Not to mention, after Microsoft, NVIDIA’s pullback from the company are signs of worry too.
This pattern traces back to the founding of OpenAI. Elong Musk has contributed between $38 million and $45 million in early funding, based on the non-profit mission of the company. As OpenAI converted in October 2026 to a for-profit public benefit corporation, Musk filed a fraud lawsuit. He, with this lawsuit, seeks $135 billion in damages, with the trial beginning April 27, 2026.
The users have even felt the impact of it all. CEO Sam Altman, after facing the backlash over retiring GPT-4o, promised “plenty of notice” before he made any future removals. Despite it, OpenAI on January 29, announced the retirement of GPT-4o. The users received just 15 days’ notice before the shutdown of GPT-4o on February 13. This move triggered 6 months of sustained consumer backlash. Not to forget, the company partnering with Department of War was a further blow over the “military company.”
In December 2025, Altman sent an internal code red memo acknowledging that the company lost control of enterprise trust and narrative. Most senior safety leadership has since exited or just lost their influence. It left OpenAI to simply navigate some critical financial and legal challenges with reduced oversight.
