Meta Platforms, under the leadership of Mark Zuckerberg, has been making significant investments in artificial intelligence, yet its market valuation does not fully reflect these efforts. The company's internal AI adoption has not reached anticipated levels across its various software offerings. Consequently, Meta is exploring the possibility of selling its surplus AI compute power to external entities as a new cloud business. This strategic consideration raises questions about the company's ability to effectively monetize its vast AI infrastructure and the long-term impact on its stock, which is currently trading below $600.
In the last quarter, Meta's capital expenditures soared to $30 billion, nearly double the previous year, with projections indicating an annual spend of $130 billion to $145 billion, primarily allocated to AI development. While a portion of this investment aims to enhance its advertising platform, the majority is directed towards advanced AI research and inference capabilities for AI models. Despite these substantial outlays, Meta's AI models have limited external applications, leading to underutilized data center capacity. Zuckerberg revealed that other AI firms have shown interest in acquiring access to Meta's compute resources at a premium. Should Meta decide to actively pursue this AI cloud business, it could potentially generate substantial revenue, drawing parallels with successful ventures by companies like SpaceX.
However, this potential shift carries inherent risks. The fact that Meta is contemplating selling its excess compute capacity suggests a challenge in identifying internal applications for its extensive AI infrastructure. Its AI chatbots currently hold a minimal market share compared to rivals such as Alphabet's Gemini, OpenAI's ChatGPT, and Anthropic's Claude. While selling compute power could provide immediate revenue, the long-term sustainability of this strategy is debatable, as AI software providers typically prefer comprehensive cloud services offered by industry leaders like Amazon Web Services, which provide a broader range of services beyond mere compute resale. Meta is still far from establishing such a comprehensive offering.
Despite the skepticism surrounding Meta's AI investments, its foundational advertising business across Facebook, Instagram, and WhatsApp continues to demonstrate robust growth. This segment witnessed a 28% year-over-year revenue increase to $61 billion in the last quarter, driven by enhanced targeting technologies. Nevertheless, this growth has not been sufficient to offset the rising expenses from AI development and the Reality Labs division, which continues to incur significant losses. As a result, Meta's operating margin declined from 43% to 31% in Q2 2026. This indicates that while the advertising arm remains strong, the escalating costs associated with AI capital expenditures are exerting pressure on the company's profitability. The market generally views these extensive AI investments with caution.
For the first time in a considerable period, Meta's operating earnings experienced a decline last quarter. Earnings before interest and taxes (EBIT) stood at $87 billion over the past year, with a projected continued decrease in the coming years if capital expenditures persist without a clear and profitable AI business model. This is largely due to the massive depreciation that will impact Meta's income statement. Although advertising revenue is expanding rapidly, it is currently insufficient to counterbalance these escalating expenses. While Meta's stock appears undervalued when compared to many of its major technology counterparts, with a price-to-earnings (P/E) ratio of 22, its earnings are expected to contract unless a significant shift occurs in its aggressive AI investment strategy or if these investments begin to generate substantial revenue swiftly. The uncertainty surrounding Meta's ability to effectively monetize its AI endeavors makes it a cautious investment at its current valuation below $600.
