OpenAI Killed Sora. That's Not a Failure Story.

OpenAI shut down the standalone Sora product on March 24, 2026. The app is gone. The API is gone. Sora.com is gone. The Sora 2 model still exists inside ChatGPT for paid subscribers, but the standalone platform that launched in September 2025 and dominated the App Store is dead. Disney's $1 billion, three-year partnership centered on Sora, first reported by Bloomberg and Variety in December 2025, is being unwound entirely. Most companies would bury this news or spin it as a "strategic pivot." OpenAI called it what it was: unsustainable economics.

This is strategy. Real strategy.

The Product Was Working

Let me be clear about what happened here. Sora didn't fail in the way products usually fail. It wasn't abandoned because nobody wanted it. The app was at the top of the App Store. Users were generating video. The model was improving. Disney bet a billion dollars on it.

The failure wasn't market adoption. The failure was the unit economics. Video generation at scale costs too much to run profitably. Each generation burns compute. Each user running multiple generations burns more compute. At App Store prices, the math doesn't work.

OpenAI faced a choice: raise prices aggressively and risk killing adoption, or keep subsidizing the product and hope efficiency gains materialized fast enough. Neither path led somewhere good.

What This Actually Means

Most leaders never get here. They don't build products successful enough to have this problem. They don't have the discipline to kill something that works because it doesn't work as a business.

OpenAI is now consolidating video generation inside ChatGPT. Sora 2 still exists. Still improves. Still serves customers who want it. But it's no longer a standalone business unit burning capital on a platform nobody was going to pay enough to sustain.

This is the "prioritize and execute" framework Ergon teaches. Identify the highest-impact problem. Focus effort there. Kill what isn't working, even if it's working well by most measures. Especially then.

The discipline to cut is rare and valuable. Most companies don't have it.

The Sora 2 Play

Here's what matters: Sora 2 is still the best video generation model available. It's integrated into ChatGPT. Every ChatGPT Plus subscriber can use it. That's a real customer base with real commitment.

The standalone app served a smaller market. Enthusiasts. Professionals. People willing to pay per-generation. The margins were never going to exist at the prices users would tolerate.

By moving Sora 2 inside ChatGPT, OpenAI solved the unit economics problem. Video generation becomes a feature in a subscription service rather than a product line with its own cost structure. The user already pays for ChatGPT. The video requests become incremental load on infrastructure you're already maintaining.

That's not killing a product. That's unbundling it into the right container.

Why This Matters More Than the Disney Deal

Disney's billion-dollar partnership is a side note in this story. Disney wanted Sora to exist as an exclusive tool. It existed. Now it doesn't. Disney gets access to Sora 2 inside ChatGPT like everyone else.

Disney is not the story. The story is that OpenAI recognized unsustainable unit economics and moved immediately.

Most leaders wait. They hope next quarter fixes the problem. They wait for a new market to emerge. They wait for AI efficiency gains that might never arrive. They call it "patience" and "long-term thinking." It's actually drift.

OpenAI didn't drift. The product was burning capital. The company cut it. That decision happened in a few months, not years.

The Lesson

This is what happens when you stop confusing success with viability. A product can be successful and unviable simultaneously. Sora was proof.

Users loved it. The technology worked. The market wanted it. None of that mattered because the company couldn't make money running it. Sustainability isn't an option. It's the foundation.

Every leader has something in their portfolio with the same profile: high usage, high burn. Most leaders don't kill it. They negotiate with themselves. They launch a premium tier. They wait for margins to improve. They hope.

OpenAI shipped the courage to fail efficiently. That's the story.