Will AI productivity lead to lower interest rates?
A version of this article originally appeared in Quartz’s Washington newsletter. Sign up here to get the latest business and economic news and insights from Washington straight to your inbox.
Last month, President Donald Trump’s chief economist mused about the artificial intelligence boom unleashing an explosion in productivity that enables one of their goals at the Federal Reserve: lower interest rates.
He outlined a scenario in which massive spending on AI data centers eventually boosts worker productivity for workers to become more productive, in turn boosting U.S. economic growth without igniting inflation. “That should take the pressure off the Fed. They should be able to lower rates,” Hassett told CNBC at the time.
This argument has an influential supporter: Kevin Warsh, Trump’s pick to be the next chair of the Federal Reserve. Last year, Warsh held up AI as “the most productivity-enhancing wave of our lifetimes — past, present and future.”
A sustained expansion in productivity is key for a long-term improvement in the standard of living. Either workers are paid more for higher output in their jobs, or companies maintain the same productivity with fewer employees. Both outcomes can keep inflation in check.
Economic experts, however, caution that it’s too early to precisely analyze the economic impact of generative AI tools since they aren’t widely used in the workforce. Adoption rates vary by sector with professional and financial services leading the pack, while manufacturing, retail and construction experience slower take-up rates.
“The data is really, really noisy,” said Martha Gimbel, the executive director of the Budget Lab at Yale University, a research group. “We shouldn’t get out over our skis on how far along humans are in implementing the technology.”
U.S. labor productivity grew at an encouraging 2% clip in recent years, higher than the pre-pandemic average and outpacing advanced European economies. But deploying AI in a way that will make Americans broadly richer in the long run still isn’t a sure bet.
There is disagreement on when AI-fueled productivity growth will start taming inflation. Former Dallas Federal Reserve President Robert Kaplan speculated that it may not be “until late 2026 or sometime in 2027” before Warsh could start convincing Fed policymakers to lower interest rates while citing AI.
On the other hand, Gimbel argues that AI’s disinflationary effect won’t be apparent until after Trump leaves the White House.
Filtering out the AI ‘noise’
AI is still in an adoption phase. Gimbel noted only 20% of firms are employing some form of AI in their operations. The share, though, is steadily growing.
A perplexing trend in the economic data has also caught the attention of policymakers. The economy expanded by 2.2% in 2025, even as overall hiring slowed to a trickle. That suggests U.S. workers are indeed becoming more productive even as fewer workers enter the labor market.
Most analysts, though, aren’t citing AI as the culprit behind the trend. More point to the staggering investments committed to AI as a reason the U.S. economy has been able to keep robust growth. Amazon, Google, Microsoft and Meta all reported spending $131 billion in the first quarter of the year on AI development, and increased their AI expenditure forecasts for the year.
It will likely take many years for AI to deliver on broad-based increases in worker output. Take electricity, which began powering commercial grids in the 1880s. In the ensuing decades, construction developers built an electric grid and transmission lines to ensure widespread access. Factory floors were then redesigned, and workers retrained to incorporate new electrical systems and procedures into their labor.
However, policymakers observe that AI isn’t something novel that popped up with the 2022 rollout of ChatGPT. What’s underway is another chapter stretching back to AI’s initial development in the mid-20th century.
“Should we expect a similar long period of access, adoption, learning, and transformation for AI? Maybe, but we are already well into it.” Federal Reserve Bank of San Francisco President Mary Daly said in a February speech. “The origins of AI date back to the 1940s and ’50s, and it has evolved much like electricity did.”
Warsh has been counseled to tune out AI hype for now. During the confirmation hearing, Sen. John Kennedy of Louisiana memorably advised Warsh to tread carefully on AI.
“Here’s my worry — that a lot of this stuff about artificial intelligence making us more productive is a bunch of hype by people who want to sell stock and an IPO,” Kennedy said. “I’d be careful there.”
Others echoed Kennedy. “I hope the Fed filters out all that noise,” Sen. Thom Tillis, another Republican member of the Senate Banking panel, told Quartz. “If they have empirical data that looks ahead and talks about the AI market expansion, capitalization of data centers… they can look at it, but it needs to be purely through an empirical lens, not through some sort of hype.”
It seems like AI is spreading like wildfire through the U.S. economy. But the people using it won’t reorganize entire companies or their whole lives overnight.
“Some of the biggest AI boosters are holding [it] to an impossible standard for a new technology,” Gimbel said. “The technology may be fast. Humans are slow.”