So, Meta Platforms (META) is going all-in on artificial intelligence. It’s not just talking about it; the company is rolling out a bunch of new initiatives at once—from shopping tools and internal reorgs to multi-million dollar content deals. It’s a classic tech company move: when in doubt, throw more AI at the problem and see what sticks.

The goal, of course, is to build out its AI ecosystem and find new ways to make money from its massive platforms. But as with any big spending push, it’s making some investors a little nervous, even as the stock tries to climb out of a recent dip.

Your New AI Shopping Assistant (Brought to You by Meta)

First up: shopping. Meta is testing a new feature inside its AI chatbot that lets you ask for product suggestions. Think of it as a personal shopper that already knows a lot about you. The tool, available to some users in the U.S. on the Meta AI web interface, responds with a carousel of products—complete with brand names, prices, and links—along with little explanations for why it picked them.

Here’s the clever part: when it can, the tool uses details Meta already has about you, like your location and inferred preferences, to tailor the suggestions. It’s a direct shot across the bow of similar features from OpenAI’s ChatGPT and Alphabet's (GOOGL) Google Gemini. Because why should they have all the fun?

Building the AI Engine Room

Behind the scenes, Meta is also shuffling the deck chairs to build a better AI engine. The company is creating a new applied AI engineering organization to support its broader “superintelligence” strategy. According to an internal memo, Maher Saba, a vice president in Meta’s Reality Labs division, will lead the group. It will report to Chief Technology Officer Andrew Bosworth and operate with what’s described as an “ultra-flat” structure—which in corporate-speak usually means fewer managers and, theoretically, faster decisions.

This new team will work with Meta’s Superintelligence Lab, led by former Scale AI CEO Alexandr Wang, to build what they’re calling a “data engine.” The idea is to improve how fast and how well AI models can be trained. They’ll focus on building AI interfaces and tools, and on generating the data, evaluations, and feedback that get fed directly back into the model development process. It’s a classic infrastructure play: if you want to win the AI race, you need the best pit crew.

Buying the News (Literally)

Then there’s the data itself. You can’t train great AI models without great data, and Meta is opening its wallet to get it. The company recently signed a deal with News Corp (NWSA) and News Corp (NWS) that could pay the media giant up to $50 million annually. The agreement, which runs for at least three years, gives Meta access to News Corp’s current and archival content from the U.S. and U.K. for AI training.

This isn’t a one-off. Meta has been signing similar agreements with several publishers, essentially funding a mini-arms race for quality training data. In the world of AI, content is still king—you just have to license it now.

What the Money People Are Saying

All this spending and activity has Wall Street’s attention. Analysts generally see a story of near-term pain for potential long-term gain.

Eric Lynch, Managing Director at Suncoast Equity, points out that concerns about rising capital expenditures—tied directly to CEO Mark Zuckerberg’s AI push—have put pressure on the stock. That’s contributed to a 4.6% decline in the share price in 2026, even though revenue grew 26% last quarter. Lynch calls the drop at a price-to-earnings ratio around 20 “largely self-inflicted” and suggests the shares could bounce back if Meta shows more discipline in its AI spending.

On the more optimistic side, Jefferies analyst Brent Thill sees the recent pullback as a potential buying opportunity. He expects Meta’s new text and image AI models, slated to launch in the first half of 2026, to change how investors view the company’s AI chops. He also highlights cost cuts in the metaverse division (Reality Labs), better AI-driven ad performance, and the huge monetization potential of WhatsApp—projecting that platform’s revenue could grow fourfold by fiscal 2029.

The bottom line? Meta is betting big that its AI investments will pay off down the road. Investors are just trying to figure out how much to pay for that bet today.

Meta Platforms shares were up 2.07% at $668.64 at the time of publication on Wednesday.