Zuckerberg Bets the Company on Personal AI Agents Reaching Billions — But Investors Aren't Buying It

Mark Zuckerberg used Meta's latest quarterly earnings call to lay out what he sees as an inevitable near-term future: one where billions of people have personal AI agents operating around the clock to manage their finances, health, relationships, and household needs. Speaking on Wednesday's investor call, he framed this not as speculation but as something he considers 'extremely unlikely' not to happen within five years. The comments represent Meta's clearest articulation yet of where the company believes the AI market is heading — and how central that trajectory is to its own commercial survival.

Zuckerberg pointed to WhatsApp as a cornerstone of this agentic future, describing it as already the leading surface where users interact with Meta AI. As more people shift toward multi-agent workflows — interacting simultaneously with several AI systems across different domains — Meta believes its messaging platforms will become increasingly indispensable infrastructure. The company has already rolled out business-facing AI agents globally on both WhatsApp and Messenger this quarter, with adoption surpassing one million businesses. Consumer-side adoption, however, remains the harder climb.

Meta is far from alone in chasing the agentic AI wave. Google has positioned custom AI agents as a centerpiece of its Search overhaul, a move that drew significant backlash from users overwhelmed by AI-generated content flooding their results. Anthropic's Claude has seen subscription growth surge, driven largely by engineers gravitating toward its agentic coding tool, Claude Code. The race to own the 'personal agent' category is intensifying across the industry, with each major player staking out a slightly different angle on what that relationship between user and AI should look like.

What separates Meta's situation from its rivals, though, is the financial pressure building in the background. The company's free cash flow plummeted 91% year over year — from $8.55 billion to just $784 million — driven heavily by AI infrastructure investments. Reality Labs, Meta's extended reality division, posted another $4.6 billion loss this quarter, bringing its cumulative losses to roughly $88 billion since 2021. Meta's stock fell almost 10% after earnings were released, signaling that investors remain skeptical about when, or whether, the company's aggressive spending will translate into sustainable returns.

Despite the turbulence, Zuckerberg is doubling down. Meta and BlackRock this week announced a joint venture to construct a $14 billion data center in El Paso, Texas, a facility that will be critical to powering the kind of always-on AI agents Zuckerberg is promising. He described personal agents as 'the foundation for our next wave of products and revenue lines,' and signaled that the company sees stronger margins in selling intelligence than in selling raw compute — even as it pursues both. The long road to billions of users starts with the million businesses already on board, but scaling from enterprise to consumer will test Meta's patience, its capital reserves, and its credibility with Wall Street.