Meta plans a new class of non-voting shares to pay for AI, acquisitions and staff pay. It gets new funding without loosening existing control, and investors who receive the shares get no say in how the money is spent.
Key takeaways
- The Verge reports that Meta will issue a new class of non-voting shares to fund its artificial intelligence push, with trading due to begin in 2027.
- According to the report, the new stock is meant for potential acquisitions and employee compensation as well as AI investment.
- Non-voting shares let a company use its equity as currency without reducing the voting power of its existing controlling shareholders.
- The size of the issue, its pricing and any conversion or sunset terms are not yet publicly known, and the overall judgement depends on them.
The argument: Meta is separating ownership from control to pay for AI
The Verge reports that Meta disclosed the new share class alongside its third-quarter earnings. The stock will carry no voting rights. It is intended to fund AI work, potential acquisitions and employee compensation, and trading is expected to start in 2027. The publication quotes the company as pointing to the scale of investment needed to compete, sustain growth and expand infrastructure.
The argument of this article is simple. The new class is mainly a governance tool presented as a funding tool. Meta could pay for AI in several ways: cash from operations, debt, or ordinary voting shares. What the new class adds is a way to issue equity, whether to staff, to sellers of acquired companies or to the market, without any change in who controls the company’s decisions. Shareholders take on the financial exposure of the AI programme, but the authority to direct it stays where it already sits.
That is not illegal and not unusual in technology. It is still a choice. As AI investment grows, companies in the sector are leaning more on their own stock to pay for it, and the question of who decides how that money is spent matters more.
Meta already has a dual-class structure, in which one smaller class of shares carries more votes per share than the publicly traded class. A third, non-voting class goes further. Future issuance can leave the existing balance of voting power where it is, rather than slowly diluting it.
Evidence: the stated uses would otherwise dilute voting power
The three purposes named in the report all involve handing out equity on a regular basis.
Employee compensation. Large technology companies pay a large share of employee pay in stock, especially to engineers and researchers. Competition for AI specialists is intense across the industry. When stock awards are made in voting shares, every grant moves a little voting power to employees, and over many years the effect adds up. Paying in non-voting shares stops that drift.
Acquisitions. Buying a company with stock instead of cash saves cash, but in a voting-share deal the sellers become voting shareholders. Using a non-voting class keeps the deal cheap in cash terms without changing the governance balance. The report says the new stock may go towards potential acquisitions. No specific deals have been identified, and it is not known whether any are being considered.
AI investment. The report presents AI infrastructure as the main reason for the change. If Meta eventually sells the new shares to raise capital, the buyers would fund the build-out but have no vote on how it is managed.
The pattern holds across all three. Each use has a cash alternative or a voting-share alternative, and in each case the new class lets the company keep the benefits of issuing equity while avoiding the main governance cost, which is diluted voting power. That is good evidence that keeping control is a central aim of the design, not a side effect.
Evidence: the technology sector has done this before
Meta is not the first large technology company to add a non-voting class. Alphabet, Google’s parent company, created non-voting Class C shares in the mid-2010s through a stock split. The rationale it gave publicly was much the same: the company wanted to keep using equity for compensation and acquisitions without weakening the control held by its existing high-vote shareholders. Snap went further and listed on the stock market with shares that carried no votes at all.
These cases show what such structures usually do in practice. In the Alphabet case, the voting and non-voting shares have since traded side by side, usually at similar prices, and the company kept issuing equity while its existing control arrangements stayed intact. The structure did what it was designed to do.
The precedent supports the main argument in two ways. It shows that non-voting classes are an established method of protecting control during long periods of equity issuance. It also shows that markets generally accept them, which helps explain why a company would expect to introduce one without much damage to its share price. Whether Meta’s version will copy Alphabet’s mechanics, for example by distributing the new shares to existing holders as a dividend or split, is not known from the material available.
Evidence: AI spending is long, uncertain and capital-hungry
The third piece of evidence is the kind of investment being financed. Building large AI systems requires data centres, specialised chips, energy supply and scarce technical staff. These costs are high, they come early, and the returns may take years. The company’s own description, as reported by The Verge, puts the emphasis on the scale of investment needed to compete and expand infrastructure.
When spending is large and returns are uncertain, shareholders tend to want more say. Investors may reasonably disagree about how fast to spend, which projects to back and when to stop. A share class with no votes takes those holders out of that discussion by design. This report does not include specific spending figures, and none are given here. The direction is clear all the same: the more capital a strategy needs, the more it matters that those supplying it have no formal voice.
This is where the governance trade-off is sharpest. A company confident in a long-term bet may see shareholder votes as a source of short-term pressure. The same feature that shields the strategy from impatience also shields it from legitimate scrutiny.
The case against: this is flexible financing, not a shift in power
There is a serious case that this reading overstates what is changing.
First, control at Meta is already concentrated through its existing dual-class structure. Public shareholders already have limited power to change company direction, so a non-voting class may make little practical difference. It stops a slow future dilution of control rather than taking power away from anyone today.
Second, nobody has to buy the shares. Investors who value votes can hold the voting class, and those who do not can hold the new one. If markets think the lack of votes matters, they can price the non-voting shares at a discount. On that view the structure is open and priced, not a hidden cost.
Third, the financing case is real. Paying staff and buying companies with stock keeps cash for capital spending and avoids taking on more debt. A company competing in a capital-intensive field has good reasons to want as many financing options as possible, and a non-voting class adds one. Calling this a governance manoeuvre may treat a practical treasury decision as a power play.
Fourth, long-term investment can be easier to sustain under stable control. Supporters of dual-class structures argue that protecting management from short-term market pressure lets companies make long-horizon bets that widely held companies might drop too early. If AI does need patient capital, a structure that guarantees patience is arguably a strength.
Fifth, the precedents point to limited harm. The non-voting classes created by other technology companies have not obviously damaged returns for the holders of those shares.
Together these points make the strongest version of the counter-argument: the new class is an incremental, disclosed and priced financing tool in a company whose control is already settled, and it serves a strategy that may benefit from stability.
The evidence that would change the conclusion
Several details not yet public would decide which reading is more accurate.
Scale of issuance. If the new class is used sparingly, mainly for compensation, the governance effect is small and the financing reading looks stronger. If it becomes the main currency for large acquisitions or big capital raises, the separation of ownership and control becomes much more significant.
Distribution mechanism. If the shares go to all existing holders in proportion, as in a split or dividend, everyone gets the same treatment at the start. If they are issued mainly to new holders, the governance gap between those funding the company and those controlling it is wider.
Conversion and sunset terms. Some multi-class structures include sunset provisions, under which enhanced or absent voting rights eventually return to one share, one vote. A clear sunset would weaken the argument made here considerably. It is not known whether any such terms apply.
Market pricing once trading begins in 2027. A persistent discount between the non-voting and voting shares would show that investors value the missing votes. Near-identical prices would support the view that the change is mainly technical.
Responses from institutional investors and index providers. Index providers and large asset managers have taken different positions on multi-class shares over time. How they treat this new class, including whether it qualifies for major indices, will show how seriously the market takes the governance change.
Use of proceeds. If capital raised through the new class goes mostly to the AI infrastructure the company has described, the funding rationale holds. If much of it goes to buybacks of voting shares, the effect would be to concentrate control further, which would strengthen the governance reading.
Until those details are known, the most defensible conclusion is the one stated here. The new class is a financing tool, but its design lets equity be spent without voting power being given up, so its most lasting effect is likely to be on governance.
Sources and further reading
- The Verge: reporting on Meta’s third-quarter earnings and the announcement of a new non-voting share class.
- Meta investor relations materials: the company’s earnings release and any filings describing the terms of the new class.
- Alphabet corporate disclosures: background on the creation of its non-voting Class C shares and the reasons given.
- Corporate governance research from academic and investor bodies: analysis of dual-class and multi-class share structures and their effects on shareholders.
Surfaced from the rss:verge signal “tech company share restructuring”. AI-assisted draft, editorially reviewed.
