
On 24 March 2026, the AI world got a surprise that few expected. OpenAI quietly pulled the plug on its much-hyped video generation product, Sora — a tool once seen as the future of filmmaking.
Just months ago, people were saying “Hollywood is finished.”
Today, the conversation has completely changed.
🚨 What exactly happened?
- Around March 24, 2026, OpenAI announced the shutdown
- The announcement came via its official X (Twitter) account
- A short message confirmed:“We’re saying goodbye to Sora.”
We’re saying goodbye to the Sora app. To everyone who created with Sora, shared it, and built community around it: thank you. What you made with Sora mattered, and we know this news is disappointing.
— Sora (@soraofficialapp) March 24, 2026
We’ll share more soon, including timelines for the app and API and details on…
This wasn’t a gradual sunset. It was abrupt — and it included:
- The consumer app
- The API access
- Most public-facing video generation features
For a product that went viral within days of launch, the shutdown raised one big question:
👉 How can something so powerful fail so quickly?
💸 The hidden problem: AI video is insanely expensive
To understand Sora’s collapse, you need to understand one thing — cost.
Here’s a simplified breakdown:
| Type | Cost per output |
|---|---|
| Text AI | $0.001 – $0.01 |
| Image AI | $0.02 – $0.10 |
| Video AI | $2 – $20+ |
A 5–10 second AI video can cost up to $20 to generate.
Why?
- A short video = 150–600 frames
- Each frame ≈ one AI-generated image
- Plus: motion consistency, physics, scene continuity
👉 In simple terms:
One video ≈ hundreds of images ≈ thousands of text responses
📉 The business didn’t make sense
Now compare cost vs revenue:
| Type | Profit Margin |
|---|---|
| Text AI | 70% – 90% |
| Image AI | 20% – 50% |
| Video AI | -50% to -300% |
That means:
👉 Companies could lose money on every video generated
This is the core reason behind the shutdown.
⚖️ Then came the legal storm
At the same time, another major problem was building up — copyright.
Major studios began pushing back:
- Disney
- Netflix
- Paramount
Their concern was simple:
“Are these models trained on our content?”
This led to:
- Legal notices
- Licensing disputes
- Risk of billion-dollar lawsuits
Even ByteDance reportedly delayed its AI video model (Seedance) after receiving pressure from studios.
💰 What about the $1 billion Disney deal?
There were reports of a $1B-level collaboration being explored between OpenAI and Disney.
But after Sora’s shutdown:
- The deal is now uncertain or stalled
- Studios became cautious due to:
- Legal risks
- Product instability
- Unclear monetization
👉 Big takeaway:
Hollywood won’t invest billions into technology that isn’t stable or legally safe.
🎥 Was AI already entering Hollywood?
Yes — but not in the way people imagined.
Platforms like Netflix and major studios were already experimenting with AI video.
However, the use cases were limited to:
- VFX support
- Background scene generation
- Pre-visualization
Not full movies.
👉 AI was becoming a tool, not a replacement.
🧠 So… is AI video over?
No. But the hype is.
The shutdown of Sora — along with delays like ByteDance’s Seedance — signals a shift:
What’s changing:
- ❌ Public hype products are slowing down
- ✅ Private, enterprise-focused development is increasing
- ⚖️ Legal frameworks are becoming necessary
- 💸 Companies are prioritizing sustainable models
🔮 What happens next?
Short term (1–3 years)
- Fewer public launches
- More closed, invite-only tools
- Stronger copyright regulations
Mid term (3–7 years)
- Cheaper compute
- Licensed datasets
- Real business models
Long term
- AI video will succeed — but as:
- A creative tool
- Not a Hollywood replacement
🧾 Final takeaway
Sora didn’t fail because the technology was bad.
It failed because:
- The cost was too high
- The legal risks were too big
- The business model wasn’t ready
AI video isn’t dead — it’s just growing up.


