Here is a contradiction worth sitting with. SpaceX just closed what is arguably one of the most expensive acquisitions of the AI cycle to buy market share in a category its target is actively losing.
Cursor's slice of the coding-tools market fell from 41% in June 2025 to about 26% by May, per spending data from Ramp. Anthropic now controls half the category.
You do not usually pay $60 billion for an asset whose defining recent trend is decline. Unless the $60 billion was never really about the asset.
Start with the room. Elon Musk reportedly told Cursor employees that Grok has fallen behind rivals, and that he isn't used to losing.
That is not a line Musk says often. He builds the thing that wins, or he says the competition is cheating, or he says the metric is wrong. Conceding the field, out loud, to a room of people he just bought, is a different register entirely.
It's the tell. The concession is the strategy, and the strategy is to purchase what he couldn't build.
The concession is the strategy, and the strategy is to purchase what he couldn't build.
Now sit with who is paying. Not xAI, the AI company that lost the coding race. SpaceX, the rocket company, is fronting the $60 billion, all-stock, representing a 3.4% dilution at its IPO valuation.
Musk merged SpaceX and xAI earlier this year, which is the technical explanation for how an aerospace balance sheet ends up underwriting an AI coding tool. But the merger doesn't make the logic clean. It makes it convenient.
The cleaner read, arguably, is that xAI, on its own, may not have been able to write this check on merit or on multiple, though the evidence does not establish this. So the rocket company's cap table appears to be spent to close a gap the AI division opened.
The gap is no mystery to anyone inside the building. Bloomberg reported that some SpaceX engineers have been slow to adopt Grok for technical work because it isn't as effective as rival tools.
More damning: within xAI itself, certain staffers have reportedly been using Anthropic's Claude to code instead of Grok. When your own AI division reaches for a competitor's product to do the work your product is supposed to do, the market-share numbers stop being abstract.
That is the real backdrop to the $60 billion. Musk reportedly spent months trying to convince businesses to use Grok for coding, while his own people were reportedly slow to adopt it.
So the question the deal is supposed to answer is whether $60 billion in stock can fix a product problem that money didn't create. Cursor's decline happened while it had every advantage — $1 billion in annualized revenue by November, explosive growth, a top-40 slot on CNBC's Disruptor 50.
It was still losing share to Anthropic through all of it. The gap between Cursor and the leaders is not a capital gap. Buying the declining side of a category does not obviously hand you the category.
There is also the matter of what SpaceX told investors, which is almost nothing. The company disclosed no details on Cursor's customer list, momentum, or revenue when it announced the deal.
Investors handed Musk a 16% one-day stock pop anyway, briefly making SpaceX the fourth most valuable company in the U.S., ahead of Amazon and Microsoft by market cap. They bid up a $60 billion acquisition without seeing the target's numbers. That is faith, not diligence.
The financing structure had teeth from the start. The April arrangement gave SpaceX the right to acquire Cursor for $60 billion or walk away — for a price. According to CNBC, the IPO filings pegged the termination cost at a $1.5 billion fee plus $8.5 billion in computing resources.
Ten billion dollars to not do the deal is a strange amount of downside to accept for a coding tool. It reads, arguably, more like a structure that could favor Musk ending up with Cursor, though the arrangement did allow SpaceX to walk away for a fee.
And there is one investor who does very well no matter how the strategic logic shakes out. Thrive Capital holds positions in both SpaceX and Cursor; the combined stake is now worth more than $10 billion.
When the same firm sits on both sides of what could be seen as a related-party-adjacent transaction, the incentive to consummate the deal at a rich price is arguably not evenly distributed among skeptics.
Then there is the wrinkle that arguably makes this among the more conflicted mega-acquisitions of the cycle, and it has nothing to do with code.
President Trump reportedly bought as much as $50,000 of SpaceX shares on June 23, two weeks after the IPO, when the stock was trading in the mid-$150 range. By Monday's close, the stock was back at its $135 IPO price — putting the president's stake underwater.
That, on its own, is a footnote. What makes it load-bearing is the surrounding relationship.
Trump and Musk are close, a brief public falling-out last summer notwithstanding — the one where Musk accused the president of sitting on the Epstein files because of how often Trump's name appears in them. They may have since reconciled, though the evidence does not describe a reconciliation event.
Meanwhile, per a Wall Street Journal analysis, SpaceX has been hoovering up a growing volume of government contracts and benefiting from the administration's deregulatory posture.
So the picture is this: a president personally long SpaceX stock, at a loss, while the company he's invested in collects growing federal business under his administration. The White House says Trump's portfolio is run by third parties tracking indexes like the Schwab 1000.
That defense is arguably thinner than it sounds, because SpaceX lobbied major indexes to change their rules for faster inclusion ahead of the IPO — which the author suggests may have contributed to the very passive ownership that now serves as the explanation, though this is the author's speculation. The company arguably built the on-ramp, then pointed to the traffic as proof of distance.
None of this touches whether Cursor is a good product. It's about the shape of the maneuver.
An AI division loses a race on merit. A rocket company with a record IPO absorbs the target using stock. A shared investor books a $10 billion mark. And a sitting president holds shares in the acquirer while, per a Wall Street Journal analysis, it collects a growing volume of federal contracts.
Every one of those pieces has a clean, corporate explanation. Together they describe capitulation dressed as ambition — a competitive position Musk failed to build, purchased with a balance sheet built for launch vehicles, insulated from scrutiny by proximity to power.
Cursor is now officially part of SpaceX, and Cursor says it will work on Grok. The stated hope is that the best coding startup, folded in, drags Grok up to the frontier.
But the numbers that justified the purchase were falling before the purchase. Anthropic owns half the category, and Cursor was shedding share to it during Cursor's best year.
The open question isn't whether SpaceX overpaid. It's whether $60 billion can reverse a gap that $60 billion didn't cause — and whether a man who isn't used to losing just paid a record sum to lose more slowly.