Meta Platforms Inc. has stunned Wall Street by raising its 2026 capital expenditure forecast to as much as $145 billion, a dramatic increase from last year's $72 billion. The announcement came during a busy earnings day that also saw Google, Amazon, and Microsoft report their quarterly results. Despite posting a 33% revenue jump its fastest quarterly growth since 2021 Meta's shares fell more than 7% in after-hours trading as investors digested the staggering spending plan. The primary driver, according to CEO Mark Zuckerberg, is the soaring cost of memory chips caused by the global AI boom. Data center buildouts have constrained supply of high-bandwidth memory (HBM) and NAND flash chips, forcing Meta to pay premium prices to secure the components essential for training and running large language models. The ripple effect has already pushed up prices for consumer electronics like laptops and smartphones, and the shortage is expected to persist for at least another year.
Meta's pivot to artificial intelligence comes after its much-hyped Metaverse venture, led by the Reality Labs division, continues to hemorrhage money. In the latest quarter, Reality Labs posted an operating loss of over $4 billion on just $402 million in revenue, bringing its total losses over the past six years to more than $80 billion. The division, which once represented Zuckerberg's grand vision for the future of digital interaction, has now been largely sidelined as Meta redirects resources toward AI. Zuckerberg acknowledged that the company had fallen behind in the AI race, particularly to rivals like Google with its Gemini models and Microsoft's partnership with OpenAI, and launched an aggressive catch-up strategy roughly 10 months ago.
That strategy involved pouring tens of billions into research and development, poaching top talent from across the industry, and hiring Alexandr Wang, founder of Scale AI, to lead the newly formed Meta Superintelligence Labs (MSL). Wang, a prominent figure in AI data infrastructure, is tasked with building a world-class research organization from scratch. Earlier this month, MSL released its first major model, Muse Spark, a proprietary AI system that the company plans to open-source in the future mirroring Meta's successful open-source strategy for previous models like Llama. Zuckerberg called the release a sign that the lab's work is on track to build a leading lab.
On the earnings call, Zuckerberg also detailed plans to launch at least two new AI agents: one for personal use and one for business applications. The personal agent is expected to help users with daily tasks such as scheduling, information retrieval, and content creation, while the business agent will focus on customer service, sales support, and internal workflows. Early testing of business AIs has reportedly seen weekly conversation volumes grow tenfold since the start of 2026, indicating strong demand from enterprise users. The agents are built on top of Muse Spark and will leverage Meta's vast social graph to provide personalized interactions.
Internally, Meta is already deploying AI at scale to improve its core products. CFO Susan Li revealed that over half a billion weekly users on Facebook and Instagram are now watching videos that have been translated and dubbed by AI in real time, breaking down language barriers and boosting global engagement. The company is also integrating Muse Spark into its recommendation systems both for content feeds and for advertising to hyper-personalize what users see. Zuckerberg emphasized that the trend over the past few years shows an increasing return on engagement and advertiser value as AI capabilities improve, and that the company is seeing clear signals that more sophisticated models lead directly to higher user retention and ad revenue.
At the same time, Meta is cutting its workforce by 10% and offering voluntary buyouts to about 7% of U.S. staff. While executives did not directly link the layoffs to automation, Li noted that a leaner operating model would help offset the substantial AI investments. The move mirrors a broader Silicon Valley trend of companies using AI to streamline operations and reduce headcount. In recent months, Google, Microsoft, and Amazon have all announced similar workforce reductions, citing the need to reallocate resources toward AI initiatives.
The memory chip crisis that is inflating Meta's costs is global in scope. Samsung and SK Hynix, the two dominant HBM producers, have been unable to keep up with demand from hyperscalers building massive AI clusters. Prices for HBM3e chips have more than doubled year-over-year, and analysts expect the shortage to persist through 2027. This has forced Meta to secure long-term supply agreements at elevated prices, contributing to its eye-watering capex forecast. The situation is further complicated by geopolitical tensions, as the majority of advanced memory production is concentrated in South Korea, with US and European expansion still years away.
Despite the spending, some experts are cautiously optimistic about Meta's AI prospects. Unlike the Metaverse a speculative bet on virtual reality that required billions in investment with uncertain returns AI technology is already generating measurable improvements in advertising revenue and user engagement. The release of Muse Spark, even as a first step, signals that Meta's research is yielding tangible results. However, the company still lags in areas like search integration and cloud AI services, where Google and Microsoft hold commanding leads. Analysts note that Meta's strength lies in its massive user base and data advantage, which could enable it to create AI products that are deeply integrated into social experiences.
Zuckerberg remains defiant, arguing that the magnitude of the investment reflects his confidence in the opportunity. He has staked Meta's future on the belief that AI will revolutionize not just social media but also how people work, shop, and communicate. With $145 billion on the line, the next few quarters will be critical in determining whether that bet pays off or becomes another costly misstep. Some investors worry that the spending could spiral out of control if the memory chip shortage continues or if AI adoption fails to meet expectations. But for now, Zuckerberg is doubling down, promising that the company will continue to invest aggressively to secure its place in the new AI-powered landscape.
Source: Gizmodo News