Meta’s AI push is showing up in ads, even as capex climbs

Meta Platforms is spending aggressively on artificial intelligence infrastructure, and the stock’s recent slump is tied to concerns that the price tag is getting ahead of returns. But new commentary built around Meta’s latest results argues the near-term payoff from AI is already feeding into its core advertising business, not just a far-off vision of more autonomous AI assistants.

In its first-quarter 2026 update, Meta reported that ad impressions delivered across its apps rose 19% year-over-year, while the average price per ad increased 12%. Those gains helped lift total revenue to $56.31 billion, up 33% year-over-year, even as the company continued to ramp costs and capital spending.

Meta also raised its full-year 2026 capital expenditure outlook to a range of $125 billion to $145 billion, citing higher component pricing and additional data center costs to support future capacity. The company reported $19.84 billion of capital expenditures in the quarter itself, a figure that indicates how quickly its AI compute build-out is accelerating.

Behind the headline ad metrics, Meta’s executives pointed to AI-driven improvements in ad performance during the quarter. In an earnings call transcript, management said enhancements to the company’s ad models delivered a 6% or higher conversion rate gain for landing page view ads. Separate updates aimed at GenAI ad creative showed that advertisers using video generation saw more than a 3% improvement in conversion rates in tests.

The same call also offered evidence that Meta’s AI is translating into distinct ad monetization products. Management said its “value optimization suite,” an AI system that helps advertisers identify higher-value customers, has an annual revenue run rate of over $20 billion, more than doubling versus the prior year. It also cited “partnership ads,” another monetization layer that management says has a run rate exceeding $10 billion.

Meta’s AI work is extending beyond ads into business messaging. Management said its “business AIs” feature, which uses AI to handle customer conversations for businesses within messaging platforms, is facilitating more than 10 million weekly conversations, up from 1 million at the start of the year. Tech coverage of the call reported that Meta is not charging for business AI tools yet for most small businesses, with leadership indicating monetization could arrive later as the product matures.

Even with improvements, the full financial picture remains mixed. Meta’s Reality Labs segment, which includes VR and other experimental hardware and software, continued to post an operating loss during the quarter, even as the Family of Apps segment generated the bulk of operating income. As for the parts of Meta’s AI roadmap that remain harder to quantify, neither Meta’s formal results materials nor the call disclosures provide a direct, dollar-for-dollar linkage between compute spending and long-term “agent” monetization, leaving investors to watch how ad pricing and conversion hold up as spending rises. What to watch next is whether Meta keeps sustaining global average price per ad gains and conversion lift while updating capex guidance, and whether business AIs move from free scale-up to a clearer charging model.