
What happened
Dan Niles said on CNBC that Meta could build a public cloud as its next big business, noting Google, Microsoft and Amazon rent out cloud computing while Meta does not.
Why it matters
If Meta rents out computing to other companies, that would give it a second way to earn from its AI spending instead of relying only on advertising, according to Niles.
What to watch
Niles still likes Meta's momentum but admits he does not like it as much as on September 7, when the shares were down about 7% for the year — the test is whether that momentum holds.
WHO IT HITSMeta shareholders and anyone tracking whether its AI spending can eventually pay off, since a public cloud would open a second revenue line next to advertising.
Summaries like this, in your inbox every morning.
Dan Niles, founder and portfolio manager of Niles Investment Management, laid out his thinking on a recent CNBC program. His firm wrote about Meta on September 7, when the shares were down about 7% for the year. Since then, Meta has launched Muse, an AI agent that sits at number one on the App Store ahead of ChatGPT — a result Niles cites as evidence the stock still has momentum, even as he concedes he does not like it as much as he did at that September 7 entry point.
The cloud idea fits a bigger theme in the article: Meta already has an enormous audience — about 3.6 billion people use at least one Meta app every day — and it pays for AI spending out of an advertising business that throws off a lot of cash. Advertising still makes up almost 98% of revenue, so a public cloud would be a genuinely new source of income if Meta chose to pursue it.
What the outcome hinges on is cost. Meta's spending plan for 2026 is so large that free cash flow is expected to fall sharply and not recover until at least fiscal 2028, and it has signed long-term contracts and data center leases that run for years. If demand or Muse sales fall short, some of that capacity could sit idle while the costs continue — making the case for a second revenue stream like cloud more pressing, but also harder to prove while the bill is still growing.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Nvidia-backed Reflection AI and French lab Mistral unveiled new open-source models this week, Beam and Le Chon…

Higgsfield updated its AI influencer tool so users can insert a custom character into existing videos, as the…

OpenAI's dots and Meta's Muse start as assistants, then use other companies' websites and services

Faraday Future plans a US robotics factory online by the end of 2026, with the first "Built in USA" robot due…

Biohub is coordinating a $1.8 billion effort to train AI models that predict cell behavior, with Meta, Google…

Microsoft's Surface Laptop Ultra, powered by Nvidia's RTX Spark Arm-based chip, starts at $2,599 and launches…
