Zuckerberg Says Billions Could Have Personal AI Agents Within Five Years
During Meta's quarterly earnings call on Wednesday, CEO Mark Zuckerberg predicted that billions of people will have personal AI agents within five years. He described systems that understand users' goals and work for them 24/7. Meta says more than one million businesses already use its agents on WhatsApp and Messenger.
Messaging could become the interface for AI agents
Zuckerberg expects personal agents to assist with finances, health, relationships and household management. WhatsApp, already Meta's leading platform for interactions with Meta AI, could become a central interface as users communicate with multiple agents.
Meta is not alone in pursuing this model. Google has made custom agents part of its Search overhaul, while demand for Anthropic's Claude has risen alongside interest in the Claude Code agentic coding assistant. For consumer products, however, usefulness must outweigh the friction and trust concerns created by an expanding AI presence.
The financial cost of Meta's ambition
Investors remain cautious: Meta shares fell almost 10% after the quarterly report. Free cash flow declined 91% year over year, from $8.55 billion to $784 million, as the company continued investing heavily in AI infrastructure.
Reality Labs lost about $4.6 billion during the quarter, taking its cumulative losses since 2021 to roughly $88 billion. Meta and BlackRock also announced a partnership to build a $14 billion data center in El Paso, Texas.
Personal agents will be “the foundation for our next wave of products and revenue lines in the months and years ahead.”
Meta expects higher margins from selling intelligence than compute, although Zuckerberg sees commercial potential in both. The plan also supports Meta's strategy for monetizing surplus AI computing capacity, with infrastructure positioned as both a cost center and a possible revenue source.
What businesses should prepare for
Companies should treat messaging-based agents as a developing customer and operations channel, not merely another chatbot format. The practical priority is to identify narrow, measurable workflows where an agent can act safely, access only necessary data and escalate decisions to people before broader consumer adoption arrives.