• Mark Zuckerberg is urging the U.S. to accelerate AI development, arguing the technology's benefits outweigh risks and warning against regulations that could slow innovation.
  • He opposes restricting open-source AI models, advocating for broad access to powerful models to boost U.S. competitiveness.
  • Meta plans to nearly double AI-related capital expenditures to $115–135 billion in 2026, signaling a massive bet on AI infrastructure.

Zuckerberg's Pro-Innovation Stance

Mark Zuckerberg has been vocal in recent interviews and a Wall Street Journal op-ed, warning that strict controls or bans on AI models, including those from China, are not an "effective solution." Instead, he argues the U.S. should focus on competing by removing bottlenecks and scaling its own AI capabilities. "The U.S. should move faster on AI, keep powerful models open-source, and avoid regulations that would slow innovation and weaken competitiveness," he said, according to people familiar with his thinking.

Meta's CEO has also publicly criticized efforts by OpenAI and Anthropic to tightly control advanced AI, advocating for "personalized superintelligence" and broad access. His comments come as Meta runs a nationwide campaign promoting an optimistic, pro-innovation narrative about AI, pushing back against "AI doomerism."

Massive AI Infrastructure Spending

Meta is doubling down on AI with plans to nearly double capital expenditures to $115–135 billion in 2026, after spending roughly $72 billion on AI infrastructure in 2025. A significant portion is earmarked for AI data centers and infrastructure. Investor reaction has been broadly supportive, with the stock rising after Meta outlined these plans and showed strong core ad-business performance.

The company has also launched its first major model, Muse Spark, from a new Superintelligence Lab, described as compact and fast but capable in science, math, and health. It is slated to integrate across Meta's apps and devices.

Open vs. Closed AI Debate

Zuckerberg's stance puts Meta at odds with rivals like OpenAI and Anthropic, which advocate for stronger guardrails and more closed control over frontier models citing safety and national security concerns. He argues that concentrating AI power in a handful of labs through strict regulation could magnify societal risks, urging a regulatory approach that preserves innovation.

His intervention in debates about limiting Chinese AI models connects directly to U.S.–China tech rivalry. He suggests the U.S. should compete by identifying bottlenecks in its own AI ecosystem rather than trying to block Chinese rivals outright.

Implications for the Industry

Meta's massive spending signals a capital injection into data-center construction, semiconductors, and cloud infrastructure, contributing to regional economic growth. The company is also partnering with infrastructure investors like BlackRock to build large data-center campuses.

Analysts note that if Meta's bet on open AI succeeds, AI assistants could become deeply embedded in daily life for billions of users. However, experts warn that large-scale deployment of powerful open-weight models could complicate efforts to mitigate misuse, necessitating new governance frameworks.

This article was updated to include details on Meta's capital expenditure plans and the launch of Muse Spark.