SpaceX disclosed in its S-1 filing that the "Spicy" and "Unhinged" modes built into its Grok chatbot could expose the company to regulatory action, litigation, advertiser revolt, and reputational harm, a remarkably candid admission for a firm chasing what may be the largest initial public offering in American history.
The Texas-based rocket and satellite company, targeting a June 12 listing on the Nasdaq, laid out the risk in Wednesday's go-public paperwork. The filing warned plainly:
"Because these modes may be more irreverent and harsher than our standard offerings, they present heightened risks."
That single sentence sits at the intersection of two enormous stories: the race to monetize artificial intelligence and the question of whether Elon Musk's corporate empire can carry the weight of a projected $1.5 trillion valuation while one of its core products keeps generating legal and political headaches abroad.
SpaceX acquired xAI, the maker of Grok, in February. In doing so, it absorbed every financial and reputational liability the AI chatbot had already accumulated, and every one still forming.
The S-1 warns that Grok's "Spicy" and "Unhinged" modes, designed to produce "less reserved" outputs, heighten the risk of regulatory scrutiny, enforcement actions, litigation, advertiser backlash, and reputational damage. The filing also acknowledged that the chatbot's image-generation and editing features had:
"enabled the creation and dissemination of nonconsensual explicit images and/or content representing women and/or children in sexualized contexts."
That is not a critic's accusation. That language comes from SpaceX's own securities filing, the document meant to persuade investors to buy shares.
As of December, SpaceX had set aside $530 million for possible litigation losses. The filing did not break out how much of that reserve is tied specifically to Grok, but the disclosure makes clear the company's lawyers see real exposure.
The legal trouble is not theoretical. Earlier this month, French prosecutors summoned Musk to Paris to face preliminary criminal charges related in part to the dissemination of child pornography and the creation of sexualized deepfake images by Grok. The exact status and forum of those charges remain unclear from the filing, but the fact that a foreign government is pursuing criminal action against the company's founder over an AI product is the kind of risk that ought to make any prospective shareholder read the fine print twice.
Apple, meanwhile, threatened to pull Grok from its App Store over complaints that the AI application was not doing enough to stop users from generating nude or overly sexualized deepfakes. When the world's largest gatekeeper to mobile distribution warns you to clean up your product, the business implications are concrete and immediate.
For a company filing for what could be the largest IPO in history, these are not footnotes. They are flashing signals that the AI arm of Musk's empire carries liabilities that no amount of rocket launches can offset.
SpaceX reported $18.6 billion in revenue in 2025, up 33 percent from a year earlier. The company has regularly launched payloads into space for NASA and transformed the launch industry with reusable rockets. None of that is in dispute.
But the financial picture is more complicated than the revenue line suggests. SpaceX posted a net loss of $4.3 billion for the three months ended March 31. Capital expenditures hit $20.7 billion. Of that, the launch and satellite businesses spent a combined $8 billion last year, while the AI business burned through $12.7 billion.
Read that again: the AI side of the house consumed more cash than the rocket side. And it is the AI side that now carries the legal baggage of a chatbot whose own maker concedes it can produce harmful content.
Observers are predicting a roughly $80 billion fundraising round. If SpaceX achieves a $1.5 trillion valuation, it would be priced at approximately 80 times its sales. A Wall Street Journal analysis noted that the top 15 U.S. companies are valued at roughly seven times their sales. SpaceX is asking investors to pay more than eleven times that multiple.
The company justifies the premium with staggering projections. The S-1 charted future revenue opportunities totaling $28.5 trillion, $26.5 trillion from AI projects, $1.6 trillion from Starlink broadband, $740 billion from Starlink mobile, and $370 billion from what the filing calls "space-enabled solutions."
"We believe we have identified the largest actionable total addressable market ('TAM') in human history."
That is a bold claim for a company that lost $4.9 billion last year. Ambition is admirable. But investors are being asked to price in a future that depends heavily on an AI product already facing criminal scrutiny in France and distribution threats from Apple.
Musk has built real businesses. SpaceX launches rockets that work. Starlink is expanding its broadband reach and generating meaningful revenue. The engineering achievements are genuine.
The question is whether the xAI acquisition, and specifically Grok's deliberate decision to offer modes that produce edgier, less filtered content, has introduced a category of risk that the market is not pricing correctly. A chatbot that can generate sexualized images of children is not a feature. It is a liability. SpaceX's own lawyers said so in a document filed with securities regulators.
The $530 million litigation reserve is a number worth watching. If French criminal proceedings advance, if Apple follows through on its threat, or if U.S. regulators begin asking their own questions, that reserve could prove insufficient. And every dollar spent defending Grok is a dollar not spent on Starship, Starlink, or the Mars ambitions that SpaceX's board has tied directly to Musk's compensation.
Conservative investors, the kind who value honest disclosure and measurable risk, should appreciate that SpaceX put the warning in the filing. Plenty of companies bury their liabilities. SpaceX named the product, named the modes, and described the harm in plain English. That is how securities disclosure is supposed to work.
But disclosure is not mitigation. Telling investors the house might catch fire does not put the fire out.
The AI gold rush has produced a familiar pattern: companies race to capture market share, cut corners on safety, and then disclose the risks only when regulators or stock exchanges force their hand. SpaceX is not unique in this regard, but its scale makes the stakes higher than most.
A $1.5 trillion valuation built partly on a chatbot that its own filing admits can produce illegal content is a bet that the upside of AI will overwhelm the downside. Maybe it will. But Musk himself has warned about the risks of getting big bets wrong when the numbers are this large.
The June 12 Nasdaq listing will test whether the market cares more about $28.5 trillion in projected opportunity or $530 million already set aside for lawsuits, and a chatbot mode that even its owner calls a heightened risk.
When a company warns you in writing that its own product might hurt you, believe it.