Meta Platforms is betting big on AI to transform its advertising business and user experience. The company's increased capital expenditures (capex) for 2026, estimated between $115 billion and $135 billion, reflect a commitment to developing "superintelligence." This investment is primarily allocated to AI infrastructure, including data center construction, AI training hardware (NVIDIA GPUs and custom silicon), and network infrastructure.
Meta's AI strategy focuses on several key areas:
1.AI-Powered Content Recommendation: Advanced algorithms learn user preferences to deliver personalized content on Facebook and Instagram, increasing user engagement and ad revenue.
2.AI Agents: CEO Mark Zuckerberg envisions AI agents that understand individual user interests and curate their entire social media experience. These agents could also create content for users and improve ad targeting.
3.Custom AI Silicon: Meta is developing its MTIA (Meta Training and Inference Accelerator) chips to reduce reliance on NVIDIA GPUs for inference workloads, improving cost efficiency.
However, investors are wary of Meta's aggressive AI spending, given the company's history with the metaverse. The Reality Labs division, responsible for metaverse projects, incurred substantial losses in 2025. The key question is whether Meta's AI investments will yield a sufficient return to justify the massive expenditures.
Analysts predict Meta's earnings will grow to $29.60 per share in 2026 and $34.39 per share in 2027. If these forecasts are accurate, Meta's stock could surge, potentially reaching a $3 trillion market cap by 2028.