Meta Platforms META
Meta is one of the largest investors in AI infrastructure and the lead open-source AI lab through its Llama model series...
AI software and platforms.
Companies building AI-native products, enterprise software, and data platforms on top of the hardware and cloud layers.
Palantir builds AI-powered data analytics and decision-support platforms for governments and enterprises. Its AIP (Artif...
Apple AAPL
Apple is deploying on-device AI across its hardware ecosystem through Apple Intelligence, powered by its own Apple Silic...
Turning models into products.
Chips and clouds are the cost side of AI. This category is about the companies that turn AI into revenue on top of that infrastructure, and the three here do it in very different ways.
Meta uses AI to rank the feeds and ads its billions of users see, and builds its own models and assistants; it planned between $125 billion and $145 billion of capital spending for 2026, among the most of any company. Palantir sells software that lets governments and companies put AI to work on their own data. Apple puts AI on more than two billion devices and, under a deal confirmed in January 2026, uses Google's Gemini models to power its revamped Siri.
How the three differ for investors.
Meta is an advertising business that funds its AI spending from ad profits; the bet is that better AI keeps users engaged and makes each ad worth more. Palantir is a software company whose results depend on winning and expanding government and enterprise contracts, and whose share price has long assumed very fast growth. Apple sells hardware and services to an installed base of more than two billion devices; for Apple, AI is a feature that has to persuade people to upgrade, and it has chosen to license a partner's models rather than lead with its own.
That makes them three very different ways to own AI applications: a profitable incumbent spending heavily, a fast-growing software vendor, and a device maker for which AI is one part of a much larger story.
What moves these stocks.
Evidence that AI raises revenue: higher ad prices and engagement at Meta, contract wins and expanding customer deals at Palantir, upgrade cycles at Apple. Against that, investors weigh what the AI costs, which for Meta means a capital spending bill that rivals the cloud companies'.
Risks to weigh.
The private labs are moving into applications themselves: OpenAI sells ads and builds a browser and devices, and Anthropic sells office tools and industry products, so they compete for the same users and budgets. Software stocks priced for fast AI growth can fall hard when growth slows. And companies that rely on another firm's models, as Apple does on Google's, depend on a competitor for a core part of the product.