Your AI provider can change the deal on you. Here's the 5-prompt audit I run to stay ready.

November 12 is the day OpenAI’s models are set to stop working directly inside Cursor. Nobody using Cursor did anything to cause that. SpaceX bought the company, OpenAI decided it couldn’t be confident in the new owner, and the model leaves.

Monday, over on YouTube, I asked which AI future you were betting on: the one you rent from a frontier lab, or the one you own.

These aren’t separate stories. Three companies just answered that question with their money.

And, if you watched today’s video, you already know the map: OpenAI is trying to own more of the AI system, NVIDIA is trying to sell the machinery that every winner needs, and Anthropic is spreading its bets across companies that compete with one another.

This article starts where today’s video stopped.

I pay more than $600 a month across three high-end AI plans. I’ll show you what each one does, what I refuse to leave inside any one provider, and the standard each plan must meet before I pay for another month. Then I’ll give you a 30-minute test that tells you whether another AI could continue one of your real projects or whether your work is trapped.

The fight is about who controls the system around the intelligence: the hardware, the product, the distribution, and the working context that makes a model useful.

The companies are placing billion-dollar bets on that question. We’re answering it ourselves every time we upload a file, save a preference, connect an account, or let one AI become the only place that understands what we’re doing.

Watching the model is the obvious move, but it’s the wrong asset. A frontier model is valuable; access to it can change overnight. The harder thing to replace is everything the model has learned about your work.

Here’s what’s inside, mapped to the article:

• The part you can’t export. The model is replaceable. Six months of accumulated working context isn’t.

• What I keep, and what I rent. The seven questions a project note has to answer before another AI can pick up the work.

• Why I pay more than $600 a month. What each of the three plans actually buys, and which job each one holds.

• How to tell if a plan pays off. The four ways a plan earns its bill, and why a good plan can’t cover an idle one’s.

• What I’d do at $20, $60, and $200-plus. Three budgets, three answers, one rule underneath.

• The 30-minute provider exit test. Run it on one real project and find out whether your work can leave.

• The Provider Independence Audit. Five prompts that run the whole test against your real setup in an afternoon, including the quarterly fire drill.

Let’s start with what each company is actually buying, because that’s what tells you where your own exposure is.