Meta selling AI computing power: admitting defeat or breaking the deadlock? Wall Street opinions are divided.

Meta's exploration of plans to sell idle computing power has sparked sharply divergent interpretations on Wall Street. Is this a pragmatic retreat after the frustration of its AI ambitions, or a strategic breakthrough that turns massive infrastructure investments into a new profit source?

According to reports, Meta is planning to sell its idle computing resources to external parties. As soon as the news broke, Meta's stock surged 9% in a single day on Wednesday, marking its best daily performance in over a year and providing a long-awaited boost to a stock that has been under pressure this year. The company declined to comment on the reports.

This move carries significant implications for investors. Meta's free cash flow is expected to record a negative value of over $1 billion in the second quarter of this year, and according to FactSet data, this metric is projected to deteriorate significantly over the next few years. If the sale of computing power materializes, it could directly improve the company's cash flow situation and provide some return on its annual capital expenditures of up to $145 billion.

Pessimistic View: Internal AI Product Growth Falls Short of Expectations

D.A. Davidson analyst Gil Luria is cautious about this. He believes that Meta's shift to selling computing power is a signal that the company is "abandoning the frontier AI" track. Since the establishment of Meta Superintelligence Labs last year, the company has launched the new Muse Spark model, but it still lags behind Anthropic and OpenAI in terms of model competitiveness.

Baird analyst Colin Sebastian also issued a warning. He noted that Zuckenberg previously hinted that the company would only consider selling computing power externally if it "overbuilt" AI infrastructure. Therefore, if Meta does launch a cloud service platform, it may mean that the scaling of its internal AI products is not progressing as expected. However, Sebastian also acknowledged that against the backdrop of Meta's increasing AI investment, building a cloud platform is itself a "rational" choice—after all, unlike peers such as Amazon and Alphabet, Meta has never sold computing resources to third parties, which has limited its profit potential and economies of scale.

Andrew Graham, founder of Jackson Square Capital, holds a similar stance. In an interview with MarketWatch, he said that Meta's decision to sell idle computing power "helps monetize its massive AI infrastructure investments and improve free cash flow expectations—which were otherwise expected to turn significantly negative in the coming years."

Optimistic View: Monetizing Idle Capacity Fuels More Investment

Jefferies analyst Brent Thill, however, offers a distinctly different assessment. In a report on Wednesday, he stated outright that criticizing the sale of computing power as "overbuilding" is "putting the cart before the horse," because market demand for computing power continues to outpace supply.

At the data level, Thill cites Jefferies' industry research data, noting that Meta's current internal infrastructure utilization rate is approximately 65%, with 35% idle capacity. Selling this part of the resources externally will directly improve overall utilization, enhance return on investment, and boost cash flow, thereby providing ammunition for further expansion of capital expenditures, rather than reflecting a contraction intention. Zuckenberg indicated in an earnings call last year that he expects Meta to be able to lease out excess computing power at "some premium."

Thill concluded that Meta is "not exiting the AI race, but turning its early aggressive capacity commitments into an option that creates strategic value."

The Cloud Computing Market Landscape Faces a New Variable

Regardless of how Wall Street interprets Meta's motives, this potential entry poses a direct impact on existing cloud computing players. After Bloomberg's report, shares of cloud computing providers CoreWeave and Nebius fell, reflecting market concerns that new competitors could compress existing market share.

Meta's next moves will largely reveal the true answer to whether this is an "offensive or defensive" move. For investors, the focus is on whether the company can successfully transform its infrastructure cost center into a revenue source while maintaining its competitive position at the AI frontier.

Risk Warning and Disclaimer

        The market carries risks, and investment must be conducted with caution. This article does not constitute personal investment advice, nor does it take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular circumstances. Investment based on this content is at your own risk.