Updated September 21, 2026. Meta stock is drawing attention as the company moves its AI strategy from model announcements toward personal agents, subscription bundles, and everyday features across its apps. The newest product catalyst is Muse, an AI agent introduced in September; Meta One, announced a week later, adds paid plans with more AI usage. Investors are weighing that potential against a much larger infrastructure bill.
Those launches help explain why investors are watching Meta (NASDAQ: META), but they do not prove that AI is already a major source of revenue. Meta’s latest results still show an advertising-led business, while the company’s AI investment is expensive and its return will take time to measure. The latest company disclosures do not attribute any single day’s share-price movement to one product launch.
Two recent launches have made Meta’s AI plans more concrete. On September 8, Meta introduced Muse, a personal AI agent that can take actions such as sending an email or booking travel, and that may keep working after a user closes the app. On September 15, Meta announced Meta One, a subscription service with expanded AI usage, creator and business tools, and additional features across Facebook, Instagram, WhatsApp, and Meta AI.
These announcements give investors two questions to assess: Can Meta turn its enormous app audience into recurring paid customers, and can it make AI improve the advertising business that already pays most of the bills? The product releases create a story to follow, but the financial results will determine whether that story becomes a durable earnings driver.
Muse is designed to act on tasks rather than only answer prompts. Meta says it can help users plan goals, open a browser, fill in forms, and coordinate work across apps. It is powered by Muse Spark, the company’s latest model, and is rolling out in the United States through iOS, Android, and the web, with support through WhatsApp. Meta says users control how much access the agent receives and that it asks for approval before actions such as sending an email or making a purchase.
For investors, the strategic appeal is distribution. Meta can introduce an AI assistant through products people already use instead of asking each person to adopt a separate platform. An agent that helps complete practical tasks could encourage more use of Meta AI and create new reasons to return to WhatsApp or other Meta apps. But adoption, reliability, safety, and user trust are not guaranteed. A product announcement is not the same as evidence that people use an agent repeatedly or that the company can monetize it profitably.
Muse also depends on a broader model effort. Meta introduced Muse Spark in April as the first model from Meta Superintelligence Labs. The company has expanded Muse Spark-powered Meta AI features across its apps, and separately introduced Muse Image, its first image-generation model from the same lab. The scale of distribution could be valuable, but investors should look for reported usage and financial results rather than assume that more AI features automatically create more revenue.
Meta One is a more direct monetization test. Meta describes individual plans with more usage of compute-intensive AI features, including image and video generation, as well as plans for creators and businesses. The company said the service launched with more than 50 features and reported 15 million subscriptions and trials to date. That figure combines subscriptions and trials; it should not be read as 15 million paying subscribers.
Meta says the basic experience across its apps and Meta AI will remain free. Paid plans are intended to sell higher limits, added creative tools, and professional features to people who want more. That could add subscription revenue and help fund costly AI services. Yet Meta has not disclosed in this announcement how much revenue the plans generate, how many trial users convert to paid plans, or how much subscribers spend on average. Those are key measurements to watch in future company reports.
The clearest current financial link is advertising. Meta’s filing says AI initiatives include content recommendations and advertising tools, and that its AI-powered discovery engine has already improved engagement and monetization of its products. That is an important distinction: AI can support the existing ad business even if consumers never pay separately for an AI assistant.
Meta’s second-quarter 2026 results offer measurable evidence about the underlying business, although they do not isolate the portion caused by AI. Revenue was $60.80 billion, up 28% year over year. Across the Family of Apps, ad impressions increased 14% and the average price per ad rose 12%. Meta reported 3.60 billion daily active people on average in June, up 3% year over year. Those numbers show a large and growing platform on which AI-driven recommendations and ad tools may build.
| Signal | Latest company-reported figure | Why it matters |
|---|---|---|
| Second-quarter revenue | $60.80 billion, up 28% year over year | Shows the overall business is growing, but does not separate AI revenue. |
| Ad delivery and pricing | Impressions up 14%; average price per ad up 12% | Advertising remains the clearest evidence of monetization. |
| Costs and operating income | Costs and expenses up 55%; operating income down 8% | Spending and one-time items can pressure profit even as sales rise. |
| 2026 capital spending outlook | $130 billion to $145 billion | Shows the scale of the infrastructure commitment investors expect to earn back. |
The expense comparison deserves context. Meta reported $42.03 billion in second-quarter costs and expenses, including $2.40 billion in legal-proceeding charges and $1.18 billion in severance expenses tied to its May 2026 headcount reduction. Those items mean the 55% increase should not be treated as a pure measure of AI spending. Still, operating income fell 8% to $18.78 billion, underscoring that rapid revenue growth does not eliminate cost pressure.
Meta’s updated 2026 capital-expenditure range is $130 billion to $145 billion, including principal payments on finance leases, to support AI efforts and the core business. The company narrowed the previous $125 billion to $145 billion range by lifting its lower bound. Investors are watching whether new data centers, servers, and networking capacity translate into stronger ad performance, paid AI use, or both. Large spending may strengthen Meta’s future products, but it also raises the hurdle for returns and can affect cash flow and depreciation expense.
First, users may try an agent without making it part of their routine. Agents also need to handle personal information and take actions reliably; privacy, security, and mistakes could limit adoption. Second, AI services consume computing capacity, so a popular feature can still have weak economics if operating costs grow faster than revenue. Third, Meta remains highly dependent on advertising. Changes in privacy rules, competition, advertiser budgets, or user engagement can affect results regardless of how capable its models become.
There is also execution risk. Meta is competing with other technology companies for AI talent, customer attention, and infrastructure. A new assistant or subscription tier could take time to reach enough users, and some users may prefer free alternatives. Meta’s filings discuss risks from regulation, competition, ad measurement, and large infrastructure investments. The company’s own outlook is therefore important alongside product announcements.
Not on its own. A new agent and subscription service can expand Meta’s opportunities, but investors still need to compare the share price with expected earnings, risks, and their own time horizon. The strongest evidence would be sustained paid adoption or measurable gains in advertising productivity without an unchecked rise in costs. Until Meta reports those outcomes, the AI developments are catalysts to monitor—not proof of a guaranteed stock return.
Sources: Meta’s Muse announcement, Meta One announcement, and Muse Spark update; Meta’s second-quarter 2026 results and Form 10-Q for the quarter ended June 30, 2026. Figures and product status reflect company disclosures available as of September 21, 2026.
Meta stock is drawing attention as Muse and Meta One turn its AI strategy into products, while $130–$145 billion in 2026 capex raises questions about returns.
Learn what Microsoft Copilot can do, compare its free and Microsoft 365 options, and follow four practical steps to use prompts, sources, Pages, and more.
A beginner-friendly guide to artificial intelligence: what AI means, how models learn, common everyday uses, key risks, and practical tips for using AI wisely.
Learn how to track nor’easter forecasts, verify travel alerts, set decision points, and prepare safe backup plans for road, air, and rail disruptions.
The 2026 Brownlow Medal count starts at 8:00 p.m. AEST on September 21. See red-carpet timing, TV and streaming options, U.S. times, and key updates.
Physint is Hideo Kojima’s new action-espionage game. Here’s what is officially confirmed about its publisher, cast, concept, development status, and release timing.
Australia’s 2026 Intergenerational Report projects the economy to 2065–66. Explore its findings on AI, productivity, aging, data centers, and digital growth.
Permanent daylight saving time is not federal law yet. Here is what a change would mean for phones, clocks, calendars, commutes, and winter mornings.
Learn how to use the BOM Weather app and bom.gov.au to check Australian forecasts, rain radar, warnings, hourly conditions, favorites, and rain alerts.
A practical guide to Netflix’s U.S. new releases for Sept. 18–24, 2026, with picks by mood, format, time commitment, and what to skip if it is not a fit.
