LinkedIn cofounder Reid Hoffman has defended the massive wave of AI infrastructure investment in the United States, arguing that spending on data centers and related projects is providing a significant boost to the broader economy and helping the country to avoid recession. According to a report by Benzinga, speaking on the Newcomer podcast, Hoffman said that billions of dollars flowing into AI infrastructure are generating economic activity far beyond technology companies. According to him, the current AI buildout is ‘the only reason we’re not in recession’, highlighting the role of a large-scale capital investments in supporting jobs and growth. He pushed back on the common perception that this investment narrowly benefits only tech companies, arguing that most of that capital doesn’t actually concentrate in the specific areas where data centers are physically built, but instead ends up more broadly distributed across the country — a dynamic he described as a clear positive.
Reid Hoffman defends AI infrastructure boom
Reid Hoffman emphasized the ripple effects this spending has on local economies, pointing to the range of workers who benefit from data center construction beyond the tech industry itself: construction workers, carpenters, real estate agents, electricians, and lawyers, among others. He argued that people tend to assume this capital infusion flows exclusively to tech companies, when in reality a massive percentage of it is actually going toward the physical work of building data centers themselves.Hoffman also said data center developers have a responsibility to work fairly with the local communities hosting these projects, framing it as essential that both developers and communities come away benefiting from the arrangement, rather than one side absorbing most of the costs or disruption.
Context: a boom not without controversy
Hoffman’s defense of data center spending comes amid a broader wave of scrutiny over how AI infrastructure investment is being financed and accounted for. Earlier this month, Meta reportedly classified some of its AI data centers as “pilot models” in order to claim federal research tax credits, a move that reduced the company’s 2025 tax bill by $3.9 billion — though its own accountants reportedly warned that the IRS could challenge that classification.Separately, Broadcom was reportedly assembling a $60 billion financing package to help Anthropic and other companies fund their AI chip and infrastructure needs, a deal that could further expand AI capacity nationwide while simultaneously boosting demand for Broadcom’s own data-center hardware.
Not everyone shares Reid Hoffman’s optimism
Hoffman’s upbeat framing stands in contrast to warnings from other prominent voices in and around the industry. AI skeptic Ed Zitron warned last month that the private credit financing fueling the AI data-center boom had become what he called a “brewing crisis,” citing rising borrowing costs and mounting construction delays as signs of real financial strain building beneath the surface.Microsoft co-founder Bill Gates has also struck a more cautious tone, suggesting the world may need to experience a serious AI-related crisis, such as a major cyberattack or similar catastrophe, before governments and the public take the technology’s risks seriously enough to respond with real urgency.