There is a financial document that reads like a distress signal.
Not the direct "we need money" kind—but the kind that says, "We have a grand narrative, a pile of ambitious projects, a grim reality, and we're masking it all with a $26.5 trillion addressable market."
SpaceX's S-1 IPO filing. The first complete financial disclosure since the xAI merger.
The Numbers Speak for Themselves
Let's start with Grok's awkward figures:
- In Q2 2026, only 0.174% of US AI users pay for Grok
- During the same period, ChatGPT's paid user ratio exceeds 6%
- Enterprise adoption: Claude surged from 21% to 48%, Gemini from 27% to 40%, while Grok edged up slightly from 4% to 7%
- US Government sector: Out of over 400 publicly disclosed AI use cases across federal agencies, only 3 mention xAI or Grok
A Reuters report used a single word to describe Grok's performance in the government sector: "flop."
Three. Out of more than 400 agencies.
The number needs no interpretation. It speaks for itself.
Contrast That With $26.5 Trillion
In that same S-1 filing, SpaceX claims to possess the "largest addressable market in human history"—with the AI segment alone estimated at $26.5 trillion.
For context:
- Gartner estimates global AI spending will reach $3.3 trillion by 2027
- Citi forecasts the global AI market could exceed $4.2 trillion by 2030
- The US nominal GDP in Q1 2026 was approximately $32 trillion
SpaceX's figure is an order of magnitude higher than any third-party forecast. It's barely less than the entire US quarterly GDP.
When your actual market share is 0.174% and your addressable market is $26.5 trillion—you're not doing market analysis. You're writing science fiction.
It's Not Just Grok
But Grok's struggles are just the tip of the iceberg. The S-1 filing also reveals other SpaceX projects still in a "very early stage of development":
- Macrohard: An agentic AI platform developed in partnership with Tesla
- Terafab: A chip manufacturing facility in collaboration with Tesla and Intel, targeting "1 terawatt of compute hardware production annually"
- Orbital Data Centers: Ultimately deploying up to 1 million satellites as orbital data centers, with potential costs exceeding $1 trillion
Three projects. All in "very early" stages. Combined, they could easily cost one or two trillion dollars.
Meanwhile, SpaceX reported a Q1 net loss of $4.3 billion, with total debt reaching $29 billion. The only profitable business unit is Starlink.
This is exactly why the IPO is so critical.
Musk Denies Involvement in Scrapping AI Executive Order Signing
Shortly after the S-1 filing was released, Trump canceled the signing ceremony for an AI safety testing executive order. Reuters reported that Musk helped "sabotage" the order.
Musk responded on X: "This is false. I don't know what was in that EO."
But I can't help but connect these two events:
With Grok's market performance so bleak and SpaceX needing an IPO to fund its ambitions, would an executive order that could slow down AI model releases be a blessing or a curse for Grok's catch-up strategy?
Maybe Musk is telling the truth. Maybe he genuinely didn't know what was in the EO.
But perhaps—just perhaps—when a company's AI business lags at 0.174%, any regulatory shift that might slow down competitors naturally draws intense attention.
My Take
Grok isn't out of chances. A 0.174% paid adoption rate implies massive room for growth—provided the product can genuinely catch up to ChatGPT, Claude, and Gemini.
But the crux of the issue is this: catching up requires time, resources, and sustained investment. SpaceX's financial situation ($4.3B Q1 loss, $29B debt) means it doesn't have infinite ammunition.
More critically, Grok's brand reputation has taken hits on multiple fronts. The "Spicy" and "Unhinged" modes may have grabbed headlines, but they've also invited regulatory scrutiny, litigation risks, and EU bans. For a company preparing for an IPO, explicitly listing these risks in the S-1 isn't a sign of confidence—it's a lawyer's mandate.
Grok's problem isn't a product issue. It's a narrative problem.
Musk has always excelled at telling grand stories: reusable rockets, Starlink, Mars colonization, autonomous driving. Each story is big enough to make people overlook present-day struggles.
But Grok's story has run into a snag: ChatGPT's story is simply too dominant.
When 6% of users have already chosen ChatGPT, when Claude's enterprise adoption has more than doubled, and when Gemini is also growing rapidly—Grok's narrative space is severely squeezed.
0.174% isn't a starting point. It's a warning: in this winner-takes-all market, the cost of being late could be far greater than anyone imagines.
Primary Sources: