Key Takeaways
- • Meta is planning a $10 billion AI data center campus launching in 2028, representing a significant escalation in the company's capital expenditure strategy amid intensifying competition for AI infrastructure dominance.
- • Meta's total AI infrastructure spending is accelerating dramatically, with the company planning to deploy $125–$145 billion in capital expenditures during 2026 alone, approximately double the prior year's $72 billion allocation.
- • The company is pursuing a geographically diversified data center strategy, with projects confirmed in Texas (BlackRock-backed facility scaling to 1 gigawatt), Louisiana (Hyperion project expanding from $10 billion to $50+ billion), and Canada ($9 billion facility in Alberta), plus India partnerships with Reliance.
- • CEO Mark Zuckerberg has anchored Meta's strategic vision on building multi-gigawatt AI clusters, including a flagship system called Prometheus, to position the company as a leader in advancing artificial superintelligence capabilities.
- • Meta's infrastructure expansion is driven by competitive necessity in the AI arms race, with Zuckerberg signaling plans for hundreds of billions in total spending on data center buildout and AI development.
Summary
Meta is committing substantial capital to an ambitious AI infrastructure expansion program, with a planned $10 billion data center campus expected to launch in 2028. This campus represents one component of the company's broader strategy to establish leadership in AI capabilities as competition intensifies across the technology sector.
The scale of Meta's commitment has expanded considerably. In 2026, the company plans to allocate between $125 billion and $145 billion toward capital expenditures, predominantly for AI infrastructure—roughly double the $72 billion deployment from 2025. This acceleration reflects CEO Mark Zuckerberg's stated commitment to advance superintelligence through the construction of multiple multi-gigawatt computing clusters, including a flagship system designated Prometheus.
Meta's geographic footprint for AI infrastructure is expanding rapidly across multiple jurisdictions. The company has initiated development of a major data center in Texas with BlackRock backing, expected to scale to 1 gigawatt capacity and come online in 2028. The Hyperion AI project in rural Louisiana, originally budgeted at $10 billion, has expanded to exceed $50 billion in total investment and continues construction through 2030. Additionally, Meta announced a $9 billion facility in Canada's Alberta province with a two to three-year construction timeline, and is pursuing India expansion through a partnership with Reliance Industries to develop a 168-megawatt AI-enabled data center at the Jamnagar campus in Gujarat.
Zuckerberg's public statements indicate the company anticipates hundreds of billions in total spending on AI data centers and related infrastructure, signaling that current-year allocations represent early-stage deployment of a multi-year capital program. These investments position Meta as a major participant in the broader AI infrastructure arms race, where dominant technology companies are competing to secure compute capacity required for advanced model training and deployment.
The scale of Meta's commitment has expanded considerably. In 2026, the company plans to allocate between $125 billion and $145 billion toward capital expenditures, predominantly for AI infrastructure—roughly double the $72 billion deployment from 2025. This acceleration reflects CEO Mark Zuckerberg's stated commitment to advance superintelligence through the construction of multiple multi-gigawatt computing clusters, including a flagship system designated Prometheus.
Meta's geographic footprint for AI infrastructure is expanding rapidly across multiple jurisdictions. The company has initiated development of a major data center in Texas with BlackRock backing, expected to scale to 1 gigawatt capacity and come online in 2028. The Hyperion AI project in rural Louisiana, originally budgeted at $10 billion, has expanded to exceed $50 billion in total investment and continues construction through 2030. Additionally, Meta announced a $9 billion facility in Canada's Alberta province with a two to three-year construction timeline, and is pursuing India expansion through a partnership with Reliance Industries to develop a 168-megawatt AI-enabled data center at the Jamnagar campus in Gujarat.
Zuckerberg's public statements indicate the company anticipates hundreds of billions in total spending on AI data centers and related infrastructure, signaling that current-year allocations represent early-stage deployment of a multi-year capital program. These investments position Meta as a major participant in the broader AI infrastructure arms race, where dominant technology companies are competing to secure compute capacity required for advanced model training and deployment.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.