Shares of Meta Platforms (NASDAQ: META) fell 3.1% in afternoon trading Friday, dragged down by a broader sell-off in AI-related technology stocks and rising investor unease over the company’s escalating capital expenditure commitments.
The decline was concentrated around investor anxiety over Meta’s projected capital expenditures for 2026, which were raised to a range of $125 billion to $145 billion.
BMO Capital characterized Meta as having the “least visible return on investment story among peers” when it comes to its artificial intelligence projects, adding further weight to the bearish sentiment.
Concerns deepened after CEO Mark Zuckerberg reportedly noted internally that AI agent development had not accelerated as expected, a candid admission that rattled confidence among investors already wary of the spending trajectory.
Macroeconomic headwinds compounded the pressure, with the Federal Reserve’s restrictive monetary stance continuing to weigh on high-growth technology names that depend on future earnings justifications.
Global antitrust scrutiny over data privacy and online age verification also added a regulatory overhang that investors were unwilling to dismiss, contributing to the broader selling pressure on the stock.
Meta shares were trading at $643.18, representing a 3.2% decline from the previous close, and the stock now sits approximately 18.6% below its 52-week high of $790 reached in August 2025.
Despite the year-to-date decline of 1.1%, the longer-term picture remains compelling for patient investors, with those who put $1,000 into Meta shares five years ago now holding a position worth approximately $1,909.
Just seven days prior, Meta shares surged 5.2% after a cluster of AI infrastructure and product announcements reframed its heavy capital spending as a path to lower costs, new revenue streams, and faster growth rather than a pure expense drag.
As part of that earlier optimism, Meta confirmed it will begin manufacturing its custom “Iris” AI chip in September in partnership with Broadcom and TSMC, a move aimed at reducing reliance on Nvidia and AMD graphics processing units.
The company is also building a $9.1 billion data center in Canada and is targeting 14 gigawatts of total compute capacity by 2027, with plans to rent excess infrastructure as an “AI Cloud” service to generate additional revenue.
Meta’s coding model, Muse Spark 1.1, is priced at roughly one-quarter the cost of comparable offerings from OpenAI and Anthropic, a deliberate strategy to attract developers into Meta’s ecosystem ahead of broader monetization.
The core risk, however, remains execution, as custom silicon, multi-gigawatt infrastructure buildouts, and aggressively priced AI models only generate returns if utilization rates, margins, and developer adoption follow through on the spending commitments.
Meta’s shares have recorded 12 moves greater than 5% over the past year, suggesting Friday’s move, while significant, does not yet reflect a fundamental shift in the market’s long-term view of the business.
