Meta’s AI spending turns a pay dispute into a boardroom fight

Pay dispute, spending problem

The Wall Street Journal described the conflict as a “power and money struggle.” Documents reviewed by the newspaper show that, in early September, three directors involved in negotiations began preparing a fallback plan in case Zuckerberg held his position. In the most extreme scenario, the group was prepared to resign from the board.

The argument over pay, the report says, grew out of the scale of Zuckerberg’s AI spending. Meta’s free cash flow fell to negative $2.12 billion in the third quarter of 2026.

Robert Pavlik, senior portfolio manager at Dakota Wealth, told Reuters that investors are concerned about rising costs, negative free cash flow and insufficient revenue from the company’s investments. Their worry is that AI could become a bottomless expense.

The bet has a financial limit

Zuckerberg has argued that aggressive AI investment is a necessary risk. At a meeting with investors in October 2025, he said Meta needed to take that risk, while acknowledging that building infrastructure on a massive scale had created a dangerous financial bubble.

In 2025, he told shareholders that Meta should expand capacity early and quickly so it would be ready for the most optimistic scenarios.

That strategy puts the board’s compensation fight in a wider context: the question is not only how much Zuckerberg should be paid, but how much financial room Meta has to keep spending before the returns arrive. Meta’s AI products are not passing safety checks and automation tests, while the crisis around Muse’s safety continues.

I think the missing piece is a clear account of what would count as evidence that this spending is working. Without one, negative cash flow and stalled product checks make “prepare for the best case” a difficult strategy to distinguish from an open-ended commitment.