The news: Meta, Google, and Microsoft are spending at historic rates in the race to secure AI dominance. Each posted record quarterly earnings last week—and warned that even higher capital expenditures are an imperative for growth, per Wired.

• Meta lifted its 2025 capital expenditure (capex) forecast to as high as $72 billion, up from $66 billion. CFO Susan Li expects “notably larger” spending next year. Quarterly revenues surged 26% to $51.2 billion, driven by AI-powered ads and Reels.
• Alphabet raised its 2025 capex projection to $93 billion—up from $75 billion earlier. Cloud revenues jumped 35% to $15.1 billion, while Gemini AI reached 650 million monthly users.
• Microsoft invested $34.9 billion this quarter alone—a 74% jump YoY—fueling its Azure and OpenAI-linked infrastructure. Revenues climbed 18% to $77 billion, but losses from OpenAI shaved off net income by $3.1 billion.

Why it’s worth watching: The companies justify their sprees as necessary to meet accelerating and future AI demand. As CEO Mark Zuckerberg put it, Meta is “front-loading capacity” in anticipation of the leap toward artificial superintelligence.

However, some analysts warn the AI boom risks becoming a Potemkin village—grand façades of progress masking fragile economics beneath.

Challenges ahead: Behind the record profits, warning signs are emerging. Investors fear the AI surge could echo previous tech bubbles—lavish capital chasing uncertain demand. Data center projects now span decades and hundreds of billions of dollars in commitments, yet monetization remains murky.

• Overbuild risk. Microsoft, Meta, and Alphabet are expanding capacity faster than enterprise adoption or regulation can keep pace. Much of that infrastructure may sit idle if AI economics don’t scale.
• Market distortion. Nvidia’s $100 billion OpenAI commitment and OpenAI’s $1.4 trillion compute plan have altered expectations, inflating supplier valuations and startup ambitions.
• Capital gravity. As Big Tech concentrates investments, smaller firms face higher compute costs and barriers to entry, consolidating AI power among a handful of players.

Economists warn that the AI race could ripple across supply chains, energy grids, and global liquidity. Every new data center draws megawatts, metals, and money, per S&P Global—fueling growth now but tightening margins later.

What it means for brands: For marketers, the AI buildout presents both an opportunity and a cautionary tale. As Big Tech chases scale, brands must chase substance—using AI not for hype, but for measurable value today, not in the future.

AI’s future isn’t guaranteed by capital—it’s earned through trust, differentiation, and adaptability. Brands that master those traits will thrive no matter how the infrastructure race unfolds.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.