The news arrived as a line in a court filing on Wednesday: Meta Platforms will not issue a new class of shares while a lawsuit over the arrangement is pending.
The stock was meant to do two jobs at once. It would have given Meta a currency for its artificial-intelligence buildout, and it would have let Mark Zuckerberg sell part of his holding without surrendering control of the company he founded. The plan was scheduled to take effect in January.
Several shareholders sued Zuckerberg and the board for breach of fiduciary duty. Their claim, as the filing describes it, is that the independent committee formed to review the transaction did not press hard enough on terms that tilted toward the founder. A trial is set for April.
Non-voting stock is an old answer to a narrow problem. Companies that want founders to keep voting control usually build a dual-class structure at the initial public offering, when outside investors buy shares that pay the same dividend and carry no vote. Meta listed that way in 2012. Zuckerberg’s Class B stock carries multiple votes per share, which leaves him with a majority of the voting power while he owns a far smaller slice of the economics.
The difficulty comes later, when a founder wants cash without losing that majority. Selling Class B shares converts them into ordinary Class A stock, and the votes travel with them. A fresh non-voting class gets around this by creating paper that cannot dilute control, then letting insiders swap high-vote shares for it.
Meta tried the maneuver once before. In 2016 the board approved a Class C reclassification that would have let Zuckerberg fund his philanthropy and keep the reins. Shareholders sued, the case reached Delaware’s Court of Chancery, and the company withdrew the proposal in 2017 before trial. What is now on hold is a descendant of that design.
Delaware makes these deals hard to defend. Where a controlling shareholder sits on both sides of a transaction, courts apply entire fairness, which shifts the burden to the company to prove a fair price and a fair process. An independent committee with genuine authority to walk away is the standard defense. The plaintiffs here argue the committee was never given the room, or the information, to bargain.
Meta’s spending plans explain why the structure was worth the trouble. The company has been adding data-center capacity at a pace its operating cash flow no longer covers with comfort, and it has told investors that capital spending will keep climbing. Debt is available, and so is borrowing against future contracts. Equity from the founder’s own holdings was a third door, and one that did not require selling new stock into the market at whatever price prevailed that week.
The pause narrows the choices rather than closing them. Meta can fund the buildout from cash, from bonds, or from the kind of private credit structures that have grown common in data-center finance. Each carries a cost the share plan was designed to avoid, whether interest, covenants or disclosure. None of them touches the control question, which is the part Zuckerberg has never been willing to trade.
That control has a commercial dimension beyond sentiment. A founder with a blocking stake can promise a ten-year buildout, refuse a break-up, and ignore quarterly pressure. Investors who buy the stock accept the arrangement because it has produced returns. The suits argue that the same arrangement, extended to a new share class, lets the founder take liquidity at a price other holders cannot influence.
For the board the calculus is about timing as much as law. Pressing ahead before the April trial would hand the plaintiffs a live transaction to attack and invite a ruling that binds future boards. Waiting costs momentum at a moment when every large technology company is competing for the same power, land and chips.
There is also the matter of what the market reads into an abandoned plan. A company that quietly sets aside a financing tool is not in distress, and Meta is not. But it now has to raise money through instruments that show up on the balance sheet and in the rating agencies’ models, at a time when investors have started asking harder questions about how the AI spending will be paid for.
Zuckerberg, for his part, has given no sign of loosening his grip. He has said over the years that the structure exists to protect a long-term strategy from short-term owners. The plaintiffs answer that protection is precisely the problem when the person protected also sets the price at which he sells.
The April trial will decide whether the committee’s work holds up. If it does, a reworked plan could return with more independent scrutiny and a longer paper trail. If it does not, Meta will have spent two years and a share class to learn that Delaware still asks the same question it asked in 2016: who was on the other side of the table, and what could they say no to.
For now the Class C shares exist only in a board resolution and a docket number. Nothing has been issued, nothing has traded, and the only public trace of the plan is a filing that says it will wait.


