Bessent Tells the Fed to Weigh AI Gains Against Inflation

Treasury Secretary Scott Bessent said on Sept. 27, 2026 that Federal Reserve policymakers should keep an "open mind" on interest rates, arguing productivity gains from artificial intelligence and deregulation will help hold US inflation down.

Treasury Secretary Scott Bessent has put a specific argument in front of the Federal Reserve: that the economy's supply side is changing fast enough to justify a less restrictive stance on interest rates. Policymakers, he said, should keep an "open mind" on where rates go from here, because productivity gains from artificial intelligence and from deregulation will help keep US inflation in check.

The remarks, reported by Bloomberg, are notable less for their novelty than for their source. A Treasury secretary commenting on the inflation outlook is routine. A Treasury secretary building a case that the Fed's inflation forecast may be too pessimistic — and framing it around technology rather than demand — is a more pointed intervention, and one the central bank's own officials will have to answer in their next round of public remarks.

The supply-side case, stripped down

The economics behind Bessent's claim are conventional even if the conclusion is contested. Inflation, in the simplest framing, is what happens when demand outruns the economy's capacity to supply goods and services. If productivity rises — if each hour of work produces more output — then firms can absorb higher wages without raising prices, and the same level of demand becomes less inflationary than it would otherwise be.

That is the mechanism Bessent is pointing to. Artificial intelligence, on this reading, is not merely a capital-spending story driving equity valuations; it is a change in how much output the existing workforce can produce. Deregulation works through a parallel channel: lower compliance costs and faster permitting reduce the friction between capital and finished output.

The difficulty is one of timing and evidence. Productivity is notoriously hard to measure in real time, and it is revised heavily after the fact. Central banks are institutionally reluctant to set policy on a forecast of a productivity boom that has not yet shown up cleanly in the data. The Fed's mandate gives it an asymmetric incentive here: being wrong about transitory inflation was expensive for its credibility, and that memory shapes how much benefit of the doubt any supply-side argument receives.

Bessent's phrasing — an "open mind" rather than an explicit demand for cuts — is calibrated to that resistance. It asks officials not to commit to a particular path, which is a lower bar than asking them to move.

Why the political framing is unavoidable

No comment from a sitting Treasury secretary about Fed rate policy arrives in a vacuum. The administration has been vocal about the level of interest rates, and the independence of the central bank has been a live subject in Washington. That context means Bessent's remarks will be read two ways at once: as an economic argument about supply, and as pressure on an institution that is meant to be insulated from it.

Both readings can be true. The productivity argument stands or falls on data, not on who makes it. But the Fed's reaction function includes a defensive component — officials are acutely aware that moving in the direction the executive branch has publicly requested invites the suggestion that they moved because it was requested. In practice, overt political advocacy can make a central bank slower to act, not faster.

For investors, that is the operative point. The signal value of Treasury commentary for near-term rate decisions is low. The signal value for the broader policy environment — fiscal posture, regulatory direction, the administration's tolerance for higher or lower nominal growth — is considerably higher.

What equity markets were pricing at the last close

Markets were not in a defensive crouch heading into these remarks. In the most recent session before Bessent spoke, all three major US benchmarks closed higher, according to licensed market data as of 20:00 GMT on Fri, 25 Sep 2026, when the last trades printed.

• The S&P 500, via the SPDR S&P 500 ETF Trust (NYSEARCA: SPY), closed at $771.35, up 0.54% from a prior close of $767.18, within a day range of $766.29 to $772.28.
• The Nasdaq 100, through the Invesco QQQ Trust (NASDAQ: QQQ), finished at $744.50, a gain of 0.46% against a prior close of $741.10 and a range of $739.64 to $745.91.
• The Dow, via the SPDR Dow Jones Industrial Average ETF (NYSEARCA: DIA), led with a 0.94% advance to $517.49 from $512.68, trading between $512.59 and $517.92.

The composition of that session is worth a note. The Dow's outperformance over the Nasdaq 100 on the day is the sort of rotation that shows up when the market is leaning toward cyclicals and rate-sensitive value rather than chasing long-duration technology. A single session proves nothing about a trend, but it sits awkwardly with the idea that the AI productivity trade alone was doing the heavy lifting.

How to test the claim rather than take it on faith

Bessent's assertion is falsifiable, which makes it more useful than most policy commentary. There are specific places where an AI-driven productivity gain would show up first, and investors can watch them without waiting for a Fed statement.

The first is unit labor costs — what it costs employers to produce a unit of output once wage growth and productivity are netted against each other. If productivity is genuinely accelerating, unit labor costs should decelerate even while nominal wages hold up. That combination is the cleanest statistical fingerprint of the story Bessent is telling.

The second is corporate margins outside the technology sector. If AI is a general-purpose productivity technology rather than a capital-spending boom concentrated among a handful of suppliers, the margin benefit should broaden into services, logistics and back-office-heavy industries. Margin expansion confined to the firms selling the technology is a different, narrower phenomenon.

The third is the breakeven inflation rate embedded in Treasury securities — the gap between nominal and inflation-protected yields, which shows what the bond market expects prices to do. Bond investors, unlike equity investors, are paid to be sceptical about inflation. If the supply-side argument is winning, it should register there before it registers in Fed language.

The near-term watch list

The next meaningful data point is not a speech but a print. Each inflation release between now and the Fed's next decision will be read against Bessent's framing, and the composition will matter more than the headline: services inflation excluding shelter is where a productivity effect would be hardest to fake.

Equally important is whether other administration officials repeat the formulation. A one-off comment is commentary; a consistent message from Treasury, the Council of Economic Advisers and the White House is a coordinated position, and markets price those differently. For now, Bessent has made an argument that the Fed is under no obligation to accept — and that the data, over the coming months, will either support or quietly bury.

Frequently asked questions

What exactly did Scott Bessent say about the Fed?

Bessent, the US Treasury Secretary, said Federal Reserve policymakers should keep an "open mind" on interest rates. His argument was that productivity gains flowing from artificial intelligence and from deregulation will help keep US inflation in check, implying the Fed's inflation outlook may be more pessimistic than the underlying supply side of the economy warrants.

How would AI productivity gains actually lower inflation?

Productivity measures output per hour worked. If it rises, firms can pay higher wages and still produce goods more cheaply per unit, which removes the pressure to raise prices. The same level of consumer demand then becomes less inflationary. The catch is that productivity is measured with a lag and revised heavily, so central banks are reluctant to set policy on a forecast of it.

Does the Treasury Secretary have any authority over Fed rate decisions?

No. The Federal Reserve sets interest rates independently through its Federal Open Market Committee. The Treasury Secretary manages government finance and debt issuance and can comment publicly, but has no vote and no formal power over monetary policy. In practice, overt political pressure can make the Fed slower to move rather than faster, to protect its independence.

Where did US stock benchmarks last close?

As of the last trade at 20:00 GMT on Friday, Sept. 25, 2026, the S&P 500 ETF (SPY) closed at $771.35, up 0.54%. The Nasdaq 100 ETF (QQQ) closed at $744.50, up 0.46%. The Dow ETF (DIA) closed at $517.49, up 0.94%, the strongest of the three benchmarks that session.

What data would confirm or disprove Bessent's argument?

Three things. Unit labor costs decelerating while nominal wages hold up would be the clearest fingerprint of rising productivity. Margin expansion spreading beyond technology firms into services and logistics would show AI is a general-purpose gain, not a narrow capital-spending boom. And bond-market breakeven inflation rates falling would show sceptical fixed-income investors are buying the story.

Why did the Dow outperform the Nasdaq 100 in the last session?

On Sept. 25, 2026, the Dow ETF gained 0.94% against 0.46% for the Nasdaq 100 ETF. That pattern typically reflects rotation toward cyclical and value-oriented companies rather than long-duration technology names. One session is not a trend, but it sits somewhat awkwardly with a narrative that the AI trade alone is driving the market higher.

Sources

• Bessent Urges Fed to Keep ‘Open Mind’ on US Inflation Outlook — Bloomberg Economics