An Advertising Empire Under Siege
For years, Meta's formula was simple: gather unparalleled data on billions of users and sell that insight to advertisers. But that fortress is facing threats. Competition from algorithmically addictive
platforms like TikTok has changed user expectations, while privacy shifts from companies like Apple have systematically weakened Meta's ability to track users across the internet. This “signal loss” made its ads less effective and more expensive, forcing a strategic crisis. In early 2026, advertisers began reporting major shifts, as Meta started overhauling its ad system, moving away from the manual controls marketers had relied on for years.
The AI Counteroffensive
Meta's answer to these threats is a massive pivot to artificial intelligence. Instead of relying on user-provided data and manual targeting, the company is building a new ad engine that runs on AI from the ground up. This has two main functions. First, AI-powered recommendation engines on Facebook and Instagram work to keep users engaged by showing them content they're likely to enjoy, mimicking TikTok's successful model. Second, a suite of AI tools, often branded under the "Advantage+" umbrella, helps advertisers create and optimize campaigns with less data. The new system, sometimes called Andromeda, uses the ad creative itself—the images, videos, and text—as the primary signal for who to target, fundamentally changing the advertising game. This allows Meta's AI to find likely customers without the granular tracking data it once depended on.
The Soaring Price of Defense
This AI-fueled defense comes with a jaw-dropping price tag. Ahead of its Q2 2026 earnings report, Meta had already told investors to brace for it, raising its full-year capital expenditure (capex) forecast to a staggering $125 billion to $145 billion. This money is pouring into building massive data centers and buying up the specialized servers and chips needed to train and run advanced AI models. The spending is so intense that analysts noted it was squeezing profit margins and could even push the company's free cash flow into negative territory for the year. The spending puts Meta in the same league as other tech giants like Microsoft and Google, but with a key difference: much of this investment is aimed at protecting its existing ad business, which still accounts for nearly 98% of its total revenue.
What the Earnings Report Reveals
Coming into Wednesday's Q2 earnings call, Wall Street wasn't just looking at revenue, which was expected to be around $60.2 billion. The real focus was on whether this enormous AI bet is starting to pay off. The stock had fallen nearly 10% year-to-date, with investors getting nervous about the scale of the spending. While the advertising business itself remains strong, with Q1 showing a 33% year-over-year revenue jump, the market is now judging Meta not just as an ad platform, but as an AI infrastructure company. Even CEO Mark Zuckerberg admitted in a July town hall that the company's AI bets "haven't come to fruition yet" and that progress has been slower than hoped. The earnings report is the first major test of whether the AI-driven ad tools are improving monetization enough to justify the cost, or if the spending is simply outrunning the returns.
