Zuckerberg says Meta’s AI bet is paying off, so why does Meta tell IRS it could fail
Meta is facing scrutiny for classifying its AI data center infrastructure as 'pilot models' to claim billions in federal research tax credits. While the company publicly touts its AI success to investors, it frames the same technology as experimental and risky in tax filings to the IRS.
Why it matters
This highlights a potential conflict between corporate financial reporting and tax optimization strategies, raising questions about the abuse of government innovation incentives.
Mark Zuckerberg has spent two years telling investors that Meta's AI spending is paying off across Instagram, WhatsApp and Facebook. When the company files its taxes, it tells the US Internal Revenue Service (IRS) a very different story. According to a New York Times report, Meta describes its AI data centers to the IRS as a large experiment that may not work.That framing lets Meta claim the Research and Experimentation Tax Credit, a break Congress created in 1981 to reward risky innovation. The payoff has been big. Meta says the credit cut its tax bill by $2 billion in 2024 and $3.9 billion in 2025, up from $700 million in 2023, the year before it started using it for data centers.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in