68% Cost Surge: Meta’s AI Spending Drives Creation of New Non-Voting Stock

Meta has confirmed the creation of a new class of non-voting capital stock, a financial instrument designed to support its aggressive expansion into artificial intelligence. The announcement, made during the company’s third-quarter earnings call, signals a strategic shift in how the tech giant manages its capital structure. By issuing shares that do not carry voting rights, Meta aims to raise funds for AI development, employee compensation, and potential acquisitions without diluting the control held by its existing Class A and B shareholders. This mechanism allows the company to maintain its founder-led structure while expanding its financial toolkit to match the scale of its technological ambitions.

The financial backdrop for this move is defined by rapid growth coupled with significant expense increases. In the most recent quarter, Meta reported revenue of $64 billion, marking a 26% increase from the same period last year. However, this top-line growth came with a steep price tag; costs and expenses rose by 68% year-over-year. Chief Financial Officer Susan Li explained that the new shares are intended to provide the flexibility needed to invest in what she described as the "unprecedented momentum" driving the company. This financial strategy is crucial for sustaining the heavy capital expenditures required to build and maintain large-scale AI infrastructure.

The primary driver of these increased costs is Meta’s intensified push into the AI sector over the past year. The company has rolled out a diverse portfolio of models and tools, including the Muse Spark and Muse Glimmer models released earlier this year. More recently, it introduced Muse Wonder and Muse Spark Open, the latter being an open-weight model with publicly available parameters that allow external developers to build upon the technology. Additionally, Meta launched Muse Code, a specialized coding agent for software developers. These releases represent a multi-faceted approach to embedding AI into both its consumer products and developer ecosystems.

User adoption metrics suggest that these efforts are gaining traction. During the earnings call, CEO Mark Zuckerberg noted that more than 10 million people in the U.S. and Canada have set up Muse personal agents in just the last month and a half. He highlighted that the product has maintained the number one position in the App Store for nearly every day of that period. Zuckerberg characterized this as evidence of clear product-market fit, stating that the company is now entering a new, more operational phase focused on scaling the service to eventually serve billions of users. This growth trajectory justifies the substantial capital outlays reflected in the company’s expense reports.

While the AI division shows promise, other parts of Meta’s business face different financial realities. Reality Labs, the division responsible for augmented and virtual reality hardware and software, continues to operate at a loss, though the deficit is narrowing. The company expects the division’s losses to shrink from 2025 levels to a range of $18 billion and $19 billion. This segment remains a significant drag on overall profitability, further emphasizing the need for diversified funding sources like the new non-voting stock to balance the books while investing in high-growth areas.

The introduction of non-voting shares also serves a governance function. By distributing this new class of stock as a dividend to existing shareholders, Meta ensures that the voting power remains concentrated among current Class A and B holders. This structure protects the existing balance of power while allowing the company to access public markets for capital raising. It is a sophisticated financial maneuver that prioritizes operational flexibility over broadening shareholder influence, a common trend among large technology firms seeking to fund long-term R&D initiatives without altering their control dynamics.

Despite the positive momentum in AI and the strategic financial adjustments, Meta is not without risk factors. Li acknowledged ongoing regulatory and legal matters that could impact the company’s financial results. Specifically, she pointed to increased scrutiny regarding youth-related issues in several markets. The company has several trials scheduled in the United States in the coming months, which could introduce additional costs or operational constraints. These legal headwinds add a layer of complexity to the company’s outlook, requiring careful management of cash flow and reserves.

The creation of this new stock class represents a significant evolution in Meta’s corporate finance strategy. It acknowledges that the era of funding AI growth solely through operating cash flow may be insufficient given the scale of investment required. By tapping into public equity markets through a non-voting instrument, Meta positions itself to continue its aggressive AI buildout. This move underscores the company’s commitment to maintaining its technological edge while navigating the financial and regulatory challenges that accompany its status as a global tech leader.

As Meta moves forward, the market will likely watch closely to see how this new capital structure affects investor sentiment and stock performance. The combination of strong revenue growth, high expense ratios, and a new funding mechanism creates a complex narrative for shareholders. The success of the Muse AI products and the continued narrowing of losses in Reality Labs will be key indicators of whether this financial strategy yields the desired long-term returns. For now, the new stock class stands as a testament to Meta’s determination to fund its AI future on its own terms.