Meta Platforms continues to invest heavily in artificial intelligence (AI), raising its capital expenditure forecast for 2026 to at least $130 billion, up from $125 billion projected earlier this year. The increase underscores the company’s commitment to expanding its AI infrastructure, particularly through building more data centers that support AI operations. This announcement came alongside Meta's second-quarter financial results, which showed revenue rising 28% year-over-year to $60.8 billion, driven primarily by its advertising business. However, costs surged 55% to $42 billion, contributing to a 14% decline in profit to $18.3 billion.

CEO Mark Zuckerberg emphasized the growing role of AI across Meta’s core businesses, noting that investments in AI are accelerating key operations. He also mentioned the potential to generate new revenue streams by selling computing power directly to other companies, a business area where Meta currently lags behind competitors like Alphabet, Amazon, and Microsoft. Meta has been in discussions with AI firms such as Anthropic, which reportedly expressed interest in purchasing up to $10 billion worth of Meta’s computing capacity.

Despite these efforts, Meta’s extensive AI spending has caused concern among investors. Following the earnings report and the raised capital expenditure guidance, Meta’s stock fell more than 8% in after-hours trading. Investor apprehension stems from the company’s growing cash burn; projections indicate that Meta could enter negative free cash flow territory in the latter half of 2026, a first since its 2012 IPO, with losses potentially deepening in the following year. Some analysts suggest capital expenditures for 2027 could reach as high as $215 to $280 billion, far surpassing current estimates.

Meta’s financial pressures are compounded by rising debt. The company’s long-term borrowings stood at $83.7 billion at the end of the second quarter, a substantial increase from near zero before 2022. Furthermore, Meta has committed to financing large-scale data center projects in Louisiana and Texas through leases and financial guarantees, adding to its off-balance-sheet liabilities. The cost of borrowing is also increasing; a recent bond issuance saw yields climb toward 7%, reflecting tightening market conditions.

While Meta’s core advertising business remains robust and benefits significantly from AI-enhanced technology, questions remain about the company’s ability to diversify. Unlike its major rivals, Meta lacks a well-established platform for leasing cloud computing services to a broad customer base. The company has launched several AI products, including an advanced AI model named Muse Spark and is exploring subscription-based premium social-media features and AI tools offered to external developers. However, some analysts view these initiatives as nascent with uncertain prospects, citing Meta’s limited presence in corporate software and concerns tied to its political and regulatory reputation.

Looking ahead, Meta’s leadership appears aware of the need to demonstrate stronger financial returns on its AI investments. CFO Susan Li indicated any significant spending moderation would likely not occur before 2028, with 2026 and 2027 focused on maximizing AI computing capacity. The company’s potential move into cloud computing by selling excess infrastructure may generate substantial revenue if successful, but competing with entrenched players in this space would be a formidable challenge.

Overall, Meta’s aggressive AI investment strategy highlights both its ambition to transform beyond a social-media company and the risks involved as it navigates elevated costs, investor skepticism, and a highly competitive technology landscape.