Jet.AI convened the most consequential shareholder vote in its short history on June 11, 2026, then postponed it. The company adjourned the special meeting to June 23 for an unusual reason: not enough shareholders had voted, even though those who did overwhelmingly approved.
The numbers explain the holdup. Of the 1,421,721 shares outstanding on the May 8 record date, holders of 486,285, about 34.2%, had cast ballots, and roughly 99% of those votes backed the transaction. The obstacle is the threshold. The deal needs a majority of all shares outstanding, so any share left unvoted counts the same as a no. For founder and executive chairman Mike Winston and chief executive George Murnane, the delay is an irritation rather than a verdict. It also stalls the final step in a reinvention they’ve spent more than a year engineering: turning a private-aviation company into a pure-play artificial intelligence infrastructure business.
What Jet.AI Shareholders Are Actually Voting On
The proposal on the ballot is a divestiture dressed as a merger. Under an amended agreement first struck in 2025, Jet.AI will spin its legacy aviation operations, its Citation and HondaJet aircraft and the customer base that comes with them, into a new entity called Jet.AI SpinCo. flyExclusive (NYSE American: FLYX), one of the larger vertically integrated private-jet operators in the United States, will then acquire that business in an all-stock transaction.
The structure is built so shareholders give up nothing on paper. They’ll keep 100% of their existing Jet.AI stock and receive flyExclusive shares for the divested aviation unit. The result is exposure to two companies where there’d been one: a stake in a scaled private-aviation operator, plus a cleaner holding in what Jet.AI intends to be a focused AI company.
The regulatory path is largely cleared. The SEC declared flyExclusive’s Form S-4 registration statement effective on April 30, 2026, and the definitive proxy went out to holders in May. Both major proxy advisory firms have lined up behind the deal: ISS and Glass Lewis each recommended that stockholders vote in favor. The shortfall is procedural. With the deal endorsed by nearly everyone who’d already voted, the board pushed the meeting two weeks rather than risk a failed count on thin turnout.
From Jet Token to a Pure-Play AI Company
The reinvention is easier to follow if you start at the beginning. Winston founded the company in 2018 as Jet Token, a private-aviation venture conceived during the blockchain wave. It built a fractional-jet and jet-card business, then layered software on top, including a consumer booking app and an AI agent for charter reservations. The aviation operations gave the company revenue and a Nasdaq listing. The software gave it a reason to keep the letters “AI” in its name.
Somewhere in building those tools, the priorities flipped. Jet.AI now describes itself as “an emerging provider of high-performance GPU infrastructure and AI cloud services,” and the flyExclusive deal is the mechanism for making that description literal. Management has been explicit about where the freed-up attention goes once the jets are gone: resources are to be deployed into high-performance GPU infrastructure and business-to-business AI cloud services. The aviation segment is being handed to an operator better suited to run it, which lets Jet.AI stop being two companies at once.
That’s the part worth keeping straight. The flyExclusive transaction is a separation, and a deliberate one. The aircraft go to a buyer whose entire business is flying them, and the company that remains keeps the data-center and computing assets that Winston now treats as the main event.
Winston’s Capital-Markets Playbook
What distinguishes this pivot from the usual mid-life strategy change is how it was financed and structured. Winston’s a CFA charterholder who spent five years running a billion-dollar merger arbitrage and event-driven book before he ever founded a company, and the reinvention bears the fingerprints of that background.
Rather than sell the aviation business outright, which would’ve forced a price in a soft market, the team used a spin-off plus an all-stock merger so shareholders capture aviation value through flyExclusive equity instead of cash. Around that core move sits a set of balance-sheet maneuvers. Jet.AI reported about $13.5 million in cash and no debt as of March 31, 2026, up from $1.8 million three months earlier, after raising roughly $19.8 million through a common-stock sale during the quarter. The board authorized a $5 million share repurchase. The company also holds a $5 million economic interest tied to SpaceX and xAI, acquired through a special-purpose vehicle, an asset that’s gained attention as SpaceX was reported to be weighing an IPO this summer.
There’s also a separate vehicle in the picture, and it shouldn’t be confused with Jet.AI itself. Winston co-sponsors and runs AI Infrastructure Acquisition Corp. (NYSE: AIIA), a SPAC hunting for AI and data-center targets; Jet.AI carries its stake in that company at roughly $17.2 million on the balance sheet. AIIA and Jet.AI are distinct public companies with different tickers, exchanges, and mandates. What links them is a single operator applying capital-markets tools (spin-offs, all-stock deals, a blank-check company) to build exposure to the same theme from more than one direction.
The Asset That’s Left: Powered Land and GPUs
Strip away the aviation business and the question becomes what Jet.AI actually sells. The answer has two parts: computing capacity and the sites to house it.
The data-center portfolio is the more developed piece. Jet.AI is advancing three North American sites that together represent more than a gigawatt of planned capacity. The anchor is a roughly 395-acre campus about 10 miles south of Winnipeg, developed through a joint venture with Consensus Core, where the partners have secured a natural-gas supply equivalent to 500 megawatts of generation along with the environmental permits to use it. Securing the gas cleared a key milestone, and the company says it’s drawn hyperscaler interest. A Maritimes project is slated for about 500 megawatts, with a letter of intent for hydro and wind power. A smaller Nevada site at Moapa, sized at more than 50 megawatts, sits on a remediated former coal-plant property and is still in power-study phase.
The development model is deliberately unglamorous. Acquire land, secure power, and pre-lease to a hyperscaler before construction starts. Winston has summarized the ambition in his own words, calling Jet.AI a “pure-play leader in powered land for AI infrastructure.” The GPU-infrastructure and AI-cloud services he points to are earlier-stage, and the legacy aviation revenue is already shrinking as the business is readied for sale.
What Rides on June 23
When the meeting reconvenes, the outcome looks less like a referendum than a turnout exercise. The board has both proxy advisers on its side and nearly all cast votes in favor. The job between now and then is mechanical: get enough of the remaining shares voted to clear a majority of those outstanding.
If it passes, Jet.AI completes a transformation that few small-cap companies attempt and fewer finish, emerging without aircraft as a company whose entire pitch is the infrastructure beneath artificial intelligence. The strategic logic is coherent, and the financial scaffolding Winston built around it is unusually elaborate for a company this size.
The harder questions come after the vote. The AI business is still, by the company’s own framing, “emerging.” The data-center sites are promising parcels with power studies and letters of intent rather than signed hyperscaler leases and operating revenue, and a company holding roughly $13.5 million in cash is making a capital-intensive bet that’ll require far more. The reinvention is nearly complete on paper. Proving the new Jet.AI can operate at the scale its name now implies is the next chapter, and it starts the moment the aviation business is gone.
