Zuckerberg Tells Investors Meta’s AI Agents Will Reach Billions of People in Five Years

On Meta’s second-quarter 2026 earnings call Wednesday, CEO Mark Zuckerberg told investors that personal AI agents built by Meta will reach billions of people within five years, calling the technology the “foundation for our next wave of products and revenue lines.” The pitch landed alongside a mixed quarter: revenue rose 28 percent year over year to $60.80 billion, but net income fell 14 percent, free cash flow collapsed 91 percent, and Meta shares dropped close to 10 percent in after-hours trading as investors weighed the promised payoff against its near-term cost.

A Five-Year Bet on Personal Agents

Zuckerberg’s most specific claim came while making the case for Meta’s AI spending. “I think that it’s extremely unlikely if you look out five years from now, for example, whatever period of time you want, that you don’t have billions of people with a personal agent that understands your goals and that is just working on your behalf 24/7 to achieve your goals in whatever the domain is that you care about,” he said, according to TechCrunch. He pointed to finances, health, relationships, and household management as example use cases.

Earlier on the call, Zuckerberg explained why Meta thinks the mainstream version of this technology looks different from what exists today. “Soon we will have agents that can work 24/7 on your behalf to help you achieve your goals and improve your life, your health, your relationships, your finances, whatever you want,” he said, according to The Verge. “The first domain that agents have really taken off in is coding. But engineers are more technical and willing to spend time making those agents work, so to build great personal agents, this needs to be a great consumer product that just works out of the box and is easy enough for billions of people to adopt and use.”

Meta is positioning WhatsApp as the delivery surface for that vision. “As we move toward a future where we’re all interacting with multiple agents, I think that WhatsApp and our other messaging surfaces are going to become increasingly important,” Zuckerberg said, per TechCrunch. WhatsApp is already the leading platform for interacting with Meta AI, the company said.

An Enterprise Pitch Beyond Advertising

Personal agents were only half of Zuckerberg’s AI pitch to investors. He also argued Meta’s enterprise AI ambitions extend well beyond the customer-service agent it launched for businesses in June. “We see a large enterprise opportunity to sell to businesses, including APIs, business agents, potentially selling compute directly, and other services that we’re building for large customers,” he said, according to TechCrunch. That strategy could open new revenue streams beyond advertising, which still drives the bulk of Meta’s business, and subscriptions.

Meta’s business agents, rolled out globally on WhatsApp and Messenger this quarter and now expanding to Instagram, have already been adopted by more than 1 million businesses, the company said. Zuckerberg described the payment model as an extension of Meta’s existing advertising relationships: “just like the ad system, effectively, we will get paid when we deliver results for those businesses.”

Zuckerberg acknowledged the shift will not come easily, calling enterprise sales a “different muscle” than the one Meta has historically used. He also said the company’s own internal coding and productivity tools, built for Meta’s own use, could eventually be sold to outside customers: “now that we have those, we feel like there’s a large opportunity to serve, whether that’s small businesses or larger businesses.”

The Compute Calculus Behind the Pitch

The clearest sign of how much Meta is betting on this strategy is its spending. The company narrowed its 2026 capital expenditure guidance to a range of $130 billion to $145 billion, raising the floor from a prior outlook of $125 billion to $145 billion, according to Meta’s own earnings release. Full-year total expenses are now projected at $165 billion to $169 billion, a range the company raised in part to account for $2.4 billion in legal-proceedings charges recognized this quarter. This week, Meta and BlackRock separately announced a partnership to build a 1-gigawatt data center campus in El Paso, Texas, reported at roughly $14 billion, part of the buildout behind those numbers.

Zuckerberg pushed back on the idea that Meta should simply resell that infrastructure for a quick return. Selling off all of its compute for a “short-term profit” would be “foolish,” he said, describing Meta’s approach instead as a “portfolio” of long-term and short-term compute commitments, even while noting Meta currently has room to sell some capacity to enterprise customers at “a significant premium over what we paid for it,” according to TechCrunch. “We believe that there will continue to be a significantly higher margin on selling intelligence rather than selling compute directly, but we think that there’s a big opportunity, obviously, to sell compute as well,” he said, according to a separate TechCrunch report. Looking further out, he tied the spending to Meta’s broader ambitions: “as we get closer to personal superintelligence, we are … going to need hardware that allows you to seamlessly interact with it.”

A Costly Quarter to Make the Case

Meta’s second quarter showed the near-term cost of that bet. Net income fell 14 percent year over year to $15.85 billion and diluted earnings per share dropped 13 percent to $6.18, even as revenue climbed 28 percent, according to the company’s earnings release. Operating margin compressed to 31 percent from 43 percent a year earlier, as total costs and expenses rose 55 percent to $42.03 billion, a figure that includes $1.18 billion in severance tied to Meta’s May 2026 layoffs of roughly 8,000 employees. Free cash flow fell to $784 million from $8.55 billion in the same quarter last year, a 91 percent drop that TechCrunch reported was exacerbated by the company’s AI infrastructure spending.

Reality Labs, Meta’s AR and VR division, lost $4.6 billion in the quarter, pushing its cumulative losses since 2021 to roughly $88 billion, TechCrunch reported. For the current quarter, Meta guided to revenue of $61 billion to $64 billion, with its tax rate expected to rise to 15 to 17 percent from a prior 13 to 16 percent.

Racing OpenAI, Anthropic, and Google on Different Ground

The Verge’s reporting suggests Zuckerberg is positioning Meta’s approach as a deliberate contrast with rivals: companies like Anthropic and OpenAI have concentrated their agent efforts on coding and enterprise work, while Google has pushed into personal AI agents with its own Gemini Spark tool. Meta’s Muse Spark AI model, which launched last quarter, recently received a 1.1 update with improved coding capabilities, and The Verge reported that the company has moved roughly 7,000 staffers onto new AI initiatives, even after laying off about 8,000 people in May.

The Verge also flagged real disadvantages in that race: Meta’s AI efforts still lag the biggest labs, and unlike Google or Microsoft, the company lacks built-in ecosystem access to users’ email or documents the way Google or Microsoft do. Meta is also navigating separate trust concerns tied to its smart glasses, a backdrop that could complicate any pitch asking people to hand sensitive parts of their lives to its agents.

Zuckerberg said Meta will share more about its personal-agent plans “soon.” Until then, the earnings call leaves investors holding a specific, quantified promise, billions of agent users within five years, against a quarter in which Meta’s profit, cash flow, and stock price all moved in the opposite direction.