Is Elon Musk’s X Money Ahead of Its Time?
Many people don’t know that one of Elon Musk’s first ventures was in the financial world at X.com, an online bank he founded in 1999.
What fewer people know is how that story ended. X.com merged with Confinity, the company became PayPal, and by the fall of 2000 the board removed Musk as CEO while he was on a plane headed to his honeymoon. He didn’t walk away from banking. He got walked out of it.
He bought the domain back in 2017. He bought Twitter in 2022. He put the two together and launched X Money in 2026.
Maybe that’s coincidence. I think it’s unfinished business.
Either way, it has roots and implications in human psychology and logic among his base of users. By offering a money account with $250,000 in FDIC coverage through its banking partner, savings rates and swipe rates that beat out the rest of the financial world, he’s built something that pays for the premium service and the AI attached to it, similar to solar power. I believe he sees the larger perspective — well after the AI craze dies off and people’s dopamine levels return to normal.
The Why
AI is no different than any other business in needing to make money to survive. The founders, the equipment, and everything in between need to be paid handsomely for their contributions to society. To simplify it further, it comes down to inputs and outputs — which in the form of money is cash received versus cash spent.
Consumers power the engine by buying premium versions of AI. That’s businesses who use AI, plus individuals who run their lives like a business.
Here’s why I think the price of that comes down.
The frontier is expensive, but “good enough” is getting cheap fast. Every model in the top tier now does the same core job — write, summarize, code, reason — and most people can’t tell you which one produced what. When the products converge, price is the only thing left to compete on. Open-weight models are a year or two behind and free, which puts a ceiling on what anyone can charge for the middle of the market. And the cost to serve a token has been falling every year since this started.
Compute is a real constraint at the very top. Chips, power, data centers — you can’t wish those into existence. But that constraint protects the frontier, not the product the average person is paying $20 a month for. Below the frontier, supply is going to run past demand.
So as artificial intelligence becomes more ubiquitous and gets treated as a human right, like the internet is nowadays, prices go down.
When something is new, the general public receives hits of dopamine from these products, because they are new. But what happens when those dopamine hits cease and the brain gets bored?
People will fall off. Profits will fall. Governments will subsidize AI for people just like the internet nowadays, at lower prices. There will be more companies doing it.
The System
Why would you pay for an AI? How do you choose between two very similar ones? Elon Musk’s rollout of X Money answers these questions.
Every time you swipe the card, you receive 3% back. These are on items you were going to purchase anyway — you’re not going out of your way or waiting to buy, which reduces friction. If you swipe your card for $10k throughout the year, you’ll receive back $300.
When that money hits your account, most people will allocate some to savings, which ranges from 1% to 3.5% nowadays. This lets them save money and earn on it while it sits there. Smart move.
Under X, you earn up to 6% APY on your balance. Using the $10k example, having $10k in your account yields you $600. That 6% is quoted as an APY, which means the compounding is already baked in. The $600 is the real number, not a number you get to compound again.
The Catch
Nothing in life is free and this is no exception.
To have access to this system, you need an X subscription. Premium+ runs $40 a month, or $395 a year if you pay annually on a web browser. That gets you the full 6% with no questions asked. You just have it.
There is another route: Premium standard at $8 a month or $84 a year. That gets you 4% instead of 6%. A watered-down version of the optimal system.
Unless.
Route $1,000 or more in qualifying direct deposits into X Money in a rolling 34-day window and you get bumped to the full 6%. Same rate as the expensive tier. The cash back doesn’t change either — 3% on eligible purchases is 3% whether you’re paying $84 or $395.
So the only lever that actually matters is the deposit.
Which tells you what X is really after. It isn’t your $84. Nobody builds a bank to collect $84. It’s your paycheck. Every piece of the fine print — the direct deposit requirement, the two-days-early access, the rate bump — is engineered to make your employer’s money land inside X instead of somewhere else. Once your paycheck lands there, your bills follow, your card follows, and you stop being a subscriber and start being a customer.
There’s a cheap door and an expensive door. The cheap door just asks you to move your paycheck.
How It Adds Up
Let’s operate under some assumptions: you swipe $10k a year and you have $10k in savings. I’ll be clear that this is a simple example. It assumes interest rates don’t move and your savings doesn’t grow — which it will, especially if you’re depositing $1k a month.
$84 is spent. You route your $1,000 to X Money. One year goes by. Now you’ve earned $300 from your swipes and $600 from your interest. That’s $900, and you did spend money to get it, so we take away $84.
This leaves you with roughly $816.
One more subtraction, because I’d rather tell you than have you find out in April. That $600 in interest is taxable as ordinary income. The $300 in cash back generally isn’t — the IRS treats a rebate on your own purchase as a discount, not income. At a 22% bracket, you’re keeping closer to $684.
Still good. Just not $816.
The Other Spectrum
Let’s say you went with Apple Savings, currently offering 3.40% interest and 1% back on most swipes.
That interest on $10k yields you $340. The 1% on $10k in swipes yields you $100. You’d now have $440. That’s roughly 46% less than the X account — or put the other way, X pays you about 85% more.
But I don’t want to win against the weak version of Apple, so let’s run the strong one. Use Apple Pay on everything and you’re getting 2% instead of 1%. That’s $200 in swipes plus $340 in interest, or $540.
X still wins by about 51%.
That’s the number I’d stand behind, because it’s the one that survives an argument.
Is Elon a Visionary?
That question has been answered by his real-life projects in Tesla, SpaceX, The Boring Company. He is a visionary, and I do believe this is a system he’s building deliberately. He is preparing for the future.
A future where the average user falls off and AI prices slow down. It’s a buffer on the future.
The fanbase on X is largely more financially savvy than those on Facebook or Instagram. He’s using their own strength to tie them into the company.
How the Bank Actually Pays for the AI
I’ve been saying the bank pays for the AI, so let me show the pipe instead of just claiming it exists.
There is no cash moving from your deposit account into a server rack. The connection runs through the subscription.
Grok is bundled into Premium and Premium+. That’s the same subscription that unlocks X Money. So the money account and the AI are sold through one line item on your statement, and only one of them has to justify the purchase.
That’s the whole trick. Every other AI company has to convince you their model is worth $20 or $40 a month, every month, forever. X doesn’t. The account earns you more than the subscription costs, so the math justifies itself and the AI comes along for free. You’re not paying for Grok. You’re getting Grok because you opened a savings account.
Now flip it to the company’s side.
An AI subscriber churns when the novelty wears off. That’s the whole problem I laid out at the top. But a banking customer doesn’t churn, because leaving means re-routing your direct deposit, moving your bills, and swapping the card in your wallet. It’s a Sunday afternoon of work to save nothing. Switching costs are the most underrated moat in business.
So X converts a subscriber who would have quit into a customer who won’t. The revenue that funds the AI stops being fragile.
And the deposits are their own business. Balances generate float. The card generates interchange on every swipe. The payment rails generate volume. That’s a real revenue line that doesn’t care whether anyone is excited about chatbots this quarter.
That’s the pipe. The AI doesn’t have to be the best. It just has to be attached to something people won’t cancel.
Where My Own Argument Gets Thin
I’d rather point at the hole than have someone else find it.
The 6% probably isn’t sustainable. The Fed’s target sits at 3.50 to 3.75%. The best high-yield savings accounts in the country top out around 4.20%. X is paying 6%, which means on most of those dollars it is paying out more than it earns. That isn’t a business model, it’s a customer acquisition cost. X’s own disclosure says the rate is accurate as of July 27, 2026 and subject to change, which is the polite way of saying enjoy it while it lasts.
So no, the system doesn’t pay for itself today. Right now it pays for you, and X is eating the difference.
Here’s why I still think the thesis holds.
The rate is the advertisement. The architecture is the product. Once your paycheck lands in X and your bills route out of it, the rate can drop to 4% and you’re not leaving. Every bank in history has run this play — teaser rate in, stickiness after. The difference is that most banks have to buy your attention. X already owns it.
Judge the design, not the promotion. The design is a payments business, a deposit business, and an AI business sold as one product to a user who only has to be convinced once.
Second thing I’ll correct, because it matters and most coverage glosses over it. X Payments LLC is not a bank. It holds no charter. Cross River Bank holds the deposits and provides the FDIC coverage. So when I say Elon built a bank, what he actually built is a front end on someone else’s charter.
I don’t think that breaks the thesis. Charters can be bought, and plenty of the fintechs you already use run the same way. But it does mean the everything app isn’t finished, and anyone telling you it is hasn’t read the footnotes.
Closing
Some companies will die and others will struggle to survive. This system is built to pay for itself over time among people with the finances to do so.
I don’t believe it’s coincidence. It’s too clean of a design to be an accident.
The Twitter acquisition may have been the side piece all along — the vehicle to tie into a bigger vision of owning a bank. Not only a bank, but a bank that provides value across social media, payments, and AI at the same time. He once said he wanted an “everything app” with a bank inside it.
The rate will come down. The subsidy will end. What’s left after that is a system where the money you already have covers the software you already use, and leaving costs more than staying.
That’s the part worth watching. Not the 6%.