Elon Musk’s SpaceX is reportedly exploring a mid-2026 IPO that could rank as the all-time largest in nominal dollars and catapult SpaceX’s valuation well into the trillion-dollar club. Add in the company’s sci-fi-to-reality approach and controversial founder, and a SpaceX IPO would be the highest-profile deal of the summer.

February brought a teaser of this mega-IPO. SpaceX acquired Musk’s cash-hungry, debt-laden artificial intelligence company, xAI, in an all-stock deal to form what could be the world’s most valuable private company. xAI developed the embattled Grok chatbot and owns the social media platform X (formerly Twitter). The combined company’s goal: space-based data centers for our literal AI overlords.

The numbers in the SpaceX-xAI tie-up are stratospheric—$1 trillion in total valuation for SpaceX, and $250 billion for xAI, although without fairness opinions from an investment bank. But that’s just more gain to tax, if the planning goes awry. Tax analysis of the SpaceX-xAI deal, plus implications for a potential dual-class SpaceX IPO, below the fold.

SpaceX’s acquisition of xAI was private (and done without consulting all investors in the companies), so all information on deal structure is necessarily indirect or speculative. (Corporate-law commentators have described the acquisition as a “fiduciary stress test.”) Musk’s communications and other reports consistently characterize the transaction as “a share exchange” through a reverse triangular merger in which xAI continued as a subsidiary of SpaceX. Reuters reported that the deal was intended to be tax-free.

This post sets out the timeline for the xAI acquisition (based on reportage), as well as tax analysis of possible deal structures. The bottom line: the xAI transaction tracks some well-established forms but likely is more complex than anyone’s reported publicly. The ultimate question, however, is not just whether the xAI acquisition was tax-free, but how the acquisition’s structure might shape SpaceX’s IPO disclosures, valuation story, and prospective treatment of former xAI shareholders.

What’s the timeline for SpaceX’s acquisition of xAI?

• In March 2025, xAI acquired X (formerly Twitter) in an all-stock transaction that valued X at $33 billion. Based on Nevada Secretary of State filings and Bloomberg Law reporting, this acquisition likely resulted in a new corporate holding company, xAI Holdco, owning two subsidiaries: the xAI Opco and the X Opco.
• In July 2025, SpaceX committed to invest $2 billion in xAI, likely for a less-than-5% stake. It’s not clear which entity—the xAI Holdco, the xAI Opco, or a special-purpose financing entity—issued equity in exchange for this investment. It’s reasonable, however, to assume that the xAI Holdco issued stock to SpaceX.
• In January 2026, Tesla invested or committed to invest $2 billion in xAI, and Humain, Saudi Arabia’s state-backed AI company, invested $3 billion in xAI. These investments may have been part of a $20 billion financing round for xAI, and each investment likely represents a less-than-5% stake in xAI. Again, this analysis assumes (reasonably, but not definitively) that the xAI Holdco issued equity in exchange for these investments. Notably, Humain’s investment occurred “just before” SpaceX acquired xAI, and Humain’s investment converted to “SpaceX shares” in the acquisition.
• In late January or early February 2026, SpaceX acquired the rest of xAI’s stock in what was reported as an all-stock reverse triangular merger. Employees of xAI could elect for xAI to redeem their shares for cash; there’s evidence that some did.
• SpaceX is considering a mid- to late-2026 IPO that could raise up to $50 billion and value the company at $1.75 trillion.

What’s the tax analysis of SpaceX’s acquisition of xAI?

Based on media reports, the SpaceX-xAI deal is intended to be tax-free. But multiple routes to this result exist—and, within these multiple routes, the reorganization requirements in § 368, the presence or absence of boot, and the step-transaction doctrine do a lot of potential work. Options and commentary follow.

Option 1: B Reorganization

The reported “share exchange” of xAI stock for SpaceX stock could qualify as a B reorganization: a tax-free acquisition of xAI’s stock “in exchange solely for . . . voting stock” of SpaceX. There are, however, two potential hitches in the application of § 368(a)(1)(B) to this transaction.

First, SpaceX may have issued some nonvoting preferred stock to some shareholders in xAI, perhaps in exchange for existing nonvoting preferred in the target corporation. This might reflect business imperatives to exchange like-for-like. In the startup and venture capital contexts, preferred stock typically carries some voting rights, at least in some circumstances. But some instruments—most notably, SAFE preferred—may not have a present right to vote. If SpaceX issued nonvoting preferred to xAI shareholders, the acquisition would not qualify as a tax-free B reorganization, full stop.

Second, there’s (notoriously) no boot in a B reorganization—that is, a valid B reorganization cannot involve consideration, other than voting stock of the acquiring corporation, paid to target corporation shareholders. In the xAI-SpaceX deal, some employee-shareholders of xAI were cashed out, perhaps for substantial sums.

Reportedly, xAI used its own money to pay these departing employee-shareholders, which means that there’s no boot under Rev. Rul. 68-285. But, as the ruling clarifies, SpaceX cannot reimburse xAI for these payments to employee-shareholders. Moreover, any integrated plan in which SpaceX’s funds are effectively used to facilitate this cash-out (including through manipulation of timing, circular cash flows, or contractual commitments) could make xAI look like a conduit for getting SpaceX’s dollars to xAI’s former employee-shareholders. Money raised in an anticipated SpaceX IPO could complicate this analysis.

Given rampant speculation that SpaceX will use its substantial free cash flow to subsidize xAI’s intensive cash needs (as much as $1 billion per month), the timing of Humain’s investment in xAI (perhaps in anticipation of the acquisition), and SpaceX’s existing periodic redemption program for employees’ stock, there’s some real risk that an xAI employee-shareholder—and it just would take one—would be treated as redeemed by SpaceX for cash.

Overall, it’s hard to be a B reorganization, and it’s unlikely that the xAI-SpaceX transaction attempted to thread this specific needle.

Option 2: Reverse Triangular A Reorganization

Alternatively, the reported “triangular merger” in which SpaceX acquired xAI could qualify as a tax-free reverse triangular reorganization under § 368(a)(2)(E) and § 368(a)(1)(A). Based on Nevada public filings, SpaceX appears to have formed a transitory merger subsidiary, which then merged with and into the xAI Holdco, with the xAI Holdco surviving. xAI shareholders presumably would have received SpaceX stock in consideration for their xAI Holdco shares. As with Option 1, there are two potential complications to tax-free treatment under this state-law structure.

First, to satisfy the requirements of § 368(a)(2)(E)(i), xAI must hold “substantially all of its [own] properties” after the acquisition. For private letter rulings, Rev. Proc. 77-37 sets a benchmark for “substantially all” at 70% of gross assets and 90% of net assets. The 90%-of-net prong tends to be more restrictive for debt-laden companies such as xAI, and any redemptions of xAI’s employee-shareholders in the acquisition could obviate Rev. Proc. 77-37’s commonly accepted standard. Essentially, what’s not boot in a B reorganization (under Rev. Rul. 68-285) can be toxic under § 368(a)(2)(E)’s “substantially all” requirement. Although Rev. Proc. 77-37 isn’t a statement of substantive law, reliance on a lower threshold could be dicey, disclosure-wise, in the context of a pending IPO.

Second, to satisfy the requirements of § 368(a)(2)(E)(ii), xAI’s shareholders must exchange “control” in the company for SpaceX voting stock. Under Treasury’s longstanding interpretation of § 368(c), “control” is defined as 80% of xAI’s combined vote and 80% of the number of shares of each nonvoting class. Where there’s potential boot (see above re: employee-shareholders) or nonvoting classes (see above re: SAFE preferred), control gets trickier and more fact-dependent—which could make the xAI acquisition a bigger point of stress when designing public disclosure for a mega-IPO.

Although reverse triangular A reorganizations offer more flexibility than B reorganizations, the xAI acquisition isn’t a clear fit under the special requirements of § 368(a)(2)(E).

Option 3: Direct A Reorganization Under Rev. Rul. 2001-46

Finally, Reuters described the xAI acquisition as “a common two-step merger process” that avoided tripping debt covenants while maintaining a limited liability shield between the SpaceX parent and xAI. This statement implicates Rev. Rul. 2001-46, which outlines a double-merger tax-free structure that’s well-established in the startup and venture capital contexts. Indeed, Nevada public filings are consistent with this type of structure for the SpaceX-xAI deal.

Under this construction of the deal, SpaceX would have acquired the xAI Holdco in a reverse triangular merger (see Option 2), then merged the xAI Holdco sideways into a newly formed Nevada limited liability company under the SpaceX parent, with that LLC surviving (and the xAI Holdco terminating). At the end of the day, the SpaceX parent would have acquired the xAI Holdco’s assets—with a limited liability shield—through an entity disregarded for tax purposes. These assets might just be equity in the xAI Holdco’s subsidiaries, namely the xAI Opco and the X Opco.

Under Rev. Rul. 2001-46, this two-step transaction could qualify as a tax-free direct merger under § 368(a)(1)(A) and the step-transaction doctrine. As a non-triangular A reorganization, the acquisition would face looser restrictions on boot (60% of total consideration) and greater flexibility to use nonvoting stock as consideration, essentially mitigating the potential issues under Options 1 and 2. This structure also alleviates pressure on a prospective SpaceX IPO—there’s less risk of pre- or post-IPO adjustments, anticipated or not, complicating the xAI acquisition’s tax-free status. Life’s just easier as a vanilla A reorganization.

From this perspective, the xAI-SpaceX deal serves as a good example of the advantages of transactions modeled on Rev. Rul. 2001-46. Indeed, direct evidence of this deal structure may emerge in SpaceX’s public disclosures in advance of its IPO.

Overlapping Ownership and § 382

Another wrinkle in SpaceX’s acquisition of xAI involves overlapping ownership. Elon Musk reportedly owns more than 50% of each of SpaceX and xAI—an entirely reasonable supposition given the companies’ histories and public statements. This overlapping ownership has implications for the usability of xAI’s net operating losses by SpaceX, if xAI has any such NOLs. (The cross-crediting of xAI’s future losses against SpaceX’s future income also may be a driver in the deal.)

Musk’s overlapping ownership may allow the acquisition to avoid § 382, which generally looks to a 50-percentage-point increase in ownership—an ownership change—to trigger restrictions on target companies’ NOL carryforwards. This analysis is fact-intensive, both in terms of whether an ownership change occurs and the quantitative usability of any NOLs after an ownership change. And, if xAI’s subsidiaries join SpaceX’s consolidated return, the separate return limitation year, or SRLY, limitations would apply in the absence of a § 382 ownership change. Still, the ability to avoid § 382 may have weighed in favor—among many other factors—of doing the deal in February rather than after a mid- to late-2026 SpaceX IPO.

What the xAI Acquisition Means for a SpaceX IPO

In the context of a potential SpaceX IPO, the tax structure of the xAI acquisition matters and may shape various non-tax aspects of the SpaceX IPO.

First, tax risk with respect to the xAI acquisition may play into securities law and regulatory aspects of the SpaceX IPO. Any uncertainty about the xAI acquisition’s tax treatment could migrate into IPO disclosure and diligence, especially if the combined company must explain the merger mechanics and the treatment of cash elections in offering materials.

In addition, any IPO-related reshuffling of capital structure or voting control may need to accommodate the tax-free status of the xAI acquisition. The issue is not just boot in the abstract, but whether any broader liquidity events for legacy xAI shareholders were effectively integrated with the acquisition. It’s generally better for evolving facts not to matter than to be forced into a position that two actions are independent for tax purposes.

Finally, the usability of xAI’s tax attributes—and the cross-crediting of ongoing xAI losses against prospective SpaceX income—may play into the projected economics of an IPO of the combined company. It seems likely that there’s a significant § 382 modeling project embedded in the IPO process, and it’s worth watching this issue in SpaceX’s disclosures.

If SpaceX proceeds with a 2026 IPO, the filing process may reveal more about the concrete mechanics of the xAI acquisition. These filings also may illuminate whether the xAI deal was simply a precursor transaction or a more central element of the company’s pre-IPO tax positioning, capitalization, and public-company narrative.