Warsh: AI Growth Signals Economic Shift, But Interest Rates Unchanged

Kevin Warsh thinks AI is a big deal. Not in a vague, tech-hype kind of way — in a we-may-need-to-rethink-how-the-entire-economy-works kind of way. Speaking at Jackson Hole on August 28, the Federal Reserve chairman called the current moment a “turning point in history,” putting AI front and center in his first major address as Fed chair.

But here’s the catch: Warsh didn’t say any of it is changing interest rates yet.

What Warsh Actually Said at Jackson Hole

The speech was dense. Warsh walked through a range of economic signals tied to AI adoption, and the picture he painted was striking. Corporate spending on equipment and intangible assets — the kind of capital outlays that include servers, software, and AI infrastructure — jumped nearly 9% over the past year. That’s the fastest pace since 2021. And more than half of that growth, per Warsh, is directly tied to AI-related buildout.

He also pointed to annualized sales figures from the leading AI labs, which he said have topped $100 billion — up more than 500% within a year. Worth noting: Warsh was careful to flag that those numbers come from external reports, not from official Federal Reserve financial statements. So take them as directional, not gospel.

The bigger conceptual move in the speech was Warsh framing AI as a potential “factor of production.” That’s a loaded term in economics. If AI genuinely qualifies, it could let the economy grow faster without automatically stoking inflation. That would be a pretty significant shift from how the Fed has had to think about growth for the past several decades. Fast growth has typically meant rising prices. AI, if it genuinely boosts productivity, could break that link — or at least bend it.

That’s the optimistic read. Warsh was careful not to sell it too hard.

The Fed’s Honest Uncertainty

He didn’t claim AI has already delivered these changes. Instead, Warsh raised a pointed question: will AI actually produce a significant and lasting productivity boost? That’s not settled. And until it is, the Fed can’t really act on it.

See also: Fed Chair Warsh Warns AI is a Historic Turning Point for Monetary Policy

Misjudging productivity gains would be costly. Keep rates too high while productivity is quietly surging, and you’re strangling an expansion unnecessarily. Cut too fast on the assumption that AI is supercharging output, and you risk overheating demand. It’s a genuinely hard call, and Warsh seemed to want everyone to understand that the Fed is sitting with that uncertainty rather than pretending it doesn’t exist.

A Fed task force is apparently working through the employment and productivity questions tied to AI. Their findings could eventually shape policy. But Warsh was clear: they’re not influencing decisions right now.

The employment angle is probably the murkiest part of all this. AI can complement workers in some roles and replace them outright in others. Which happens more depends on the industry, the speed of adoption, and the specific skills involved. Warsh raised it as a concern without offering a resolution, because there isn’t one yet. No details on which sectors the task force is focusing on. Unclear when their work wraps up.

Who Actually Benefits — and Who Might Not

One of the more sobering threads in Warsh’s speech was the distribution question. Even if AI drives massive economic gains, those gains won’t necessarily spread evenly. Owners of scarce resources — advanced chips, data centers, proprietary models — are positioned to capture a disproportionate share of the new wealth. Companies that successfully deploy AI could cut costs and scale output. Consumers might see better or cheaper services. But workers in affected industries? That’s murkier.

Warsh didn’t take a side on any of it. He laid out the scenarios and basically said the Fed is watching.

See also: XRP Plummets to $1.38 as Fed Chair Warsh’s 54% Inflation Warning Shakes Markets

What the Fed is specifically watching, it seems, is the “second derivative” of AI investment. Not just whether spending is going up — it clearly is — but whether the rate of increase is speeding up or slowing down. A deceleration in that growth rate could mean companies are quietly revising their expectations about AI’s profitability. That would matter a lot for economic planning, and probably for rate decisions down the line.

So the Fed’s position right now is essentially: AI is probably important, maybe transformational, but we don’t have the hard data yet to know how it affects growth, jobs, or prices in a way that changes what we do with rates. Warsh spent a lot of words making that case carefully.

Corporate AI infrastructure spending hit its highest growth rate since 2021, and annualized AI lab revenues crossed $100 billion.