Central Bank AI: The Fed Opens Its Doors to Big Tech

The Precedent: Greenspan and the Productivity Bet

In 1996, Federal Reserve Chairman Alan Greenspan advanced a thesis that ran against internal consensus. American productivity was accelerating thanks to computing.

The mechanism was straightforward: more technology, more output per hour worked, downward pressure on prices. Greenspan kept rates low as unemployment fell toward 4%.

Staff economists predicted a return of inflation. Instead, sustained growth arrived. The reading of a general-purpose technology became, in effect, monetary policy.

Today the context differs. The structure remains identical. A central bank tries to read a technological revolution in real time, with all the risks of a misreading baked in.

What the Fed Actually Did

The Federal Reserve established the Chairman's Task Forces for Advancing Monetary Policy. One of them is named Productivity and Jobs.

The mandate is explicit: assess the economic impact of new general-purpose technologies, including artificial intelligence, to guide decisions on employment and inflation. Chairman Kevin Warsh asked leaders to measure "the pace, scope, and economic impact" of these technologies. The official page carries the date of July 9, 2026, as documented on the Board of Governors website^[1].

Warsh now leads the Fed — a politically significant detail in its own right. His emphasis on productivity echoes the tradition that views technology as a disinflationary engine.

This is an institutional change, not an academic exercise. A central bank is formally integrating AI into its decision-making framework, with named leaders and a written mandate. The central bank AI topic has moved from research papers to governance.

The Composition Matters More Than the Mandate

The names reveal the true architecture of power here.

At the helm are Marc Andreessen, co-founder of Andreessen Horowitz, Charles I. Jones, economics professor at Stanford and on leave at Anthropic, and Asha Sharma, Executive Vice President and CEO of Xbox at Microsoft. Three profiles directly tied to the capital and infrastructure of artificial intelligence.

Note the origins: venture capital, a frontier lab, a hyperscaler.

The Fed has chosen to listen to those who build and finance technology, beyond traditional academic economists. This broadens technical expertise. It also introduces a channel of private influence inside the most important monetary institution on the planet.

The foundational position remains valid: the AI race is a problem of access to semiconductors, as much as to models. Whoever controls the fabs and the compute controls the outcome. The presence of a hyperscaler at the Fed table confirms where real power resides today.

The Mechanism: Why This Shifts Monetary Policy

Productivity determines the neutral interest rate. It raises the growth potential and changes the level of rates compatible with price stability.

Here is the point: estimates of future productivity become a direct input into rate decisions. Whoever defines that estimate shapes the global cost of money.

If the task force concludes that AI accelerates productivity, the Fed gains room to tolerate tight labor markets and rapid growth. The outcome resembles the late Greenspan years: lower rates than historical models would suggest.

Private capital thereby obtains a lever over the assumption underpinning the entire yield curve.

The Capture Risk, With a Precedent

History offers a precise warning. In 1971, Fed Chairman Arthur Burns yielded to pressure from the Nixon administration and maintained an accommodative policy heading into the 1972 elections. The result was the inflation of the 1970s, with the rate exceeding 12% in 1974.

The lesson concerns the capture of the institution by outside interests.

Today the source of pressure changes in nature. Tech capital brings genuine expertise and, at the same time, a direct economic interest in low rates and high valuations. An adviser who holds stakes in AI labs benefits from an optimistic reading of productivity.

This generates a structural conflict, not merely individual suspicion. The Fed will need to build robust separation walls. The quality of those walls will determine the credibility of the entire exercise.

My View: The Line Between the Fed and Big Tech Is Thinning

Here is the thesis in one line: the entry of Big Tech leaders into the Fed's advisory bodies marks the beginning of a structural merger between tech capital and monetary policy.

This merger has partial precedents. Finance already sat at the Fed's tables through primary dealers and regional boards. The novelty lies in the direct entry of those who own the compute factories and frontier models.

I acknowledge the risk in my reading.

I would change my mind in the face of two facts: a task force mandate limited to purely public academic research, and the absence of operational recommendations in FOMC communications within eighteen months. In that case it would remain a symbolic exercise. The facts so far point in the opposite direction.

Three Implications for Capital

The structure generates concrete consequences for those allocating capital now. The productivity thesis, once institutionalized, moves from speeches to official economic projections. From there it enters prices.

1. Family Offices and Sovereign Wealth Funds (36-month horizon): overweight assets tied to AI infrastructure and semiconductors, as the Fed is institutionalizing the productivity thesis.
2. Chief Risk Officers (18-month horizon): add to VAR models a scenario of a neutral rate revised upward due to AI — currently absent from historical calibrations.
3. CFOs and Investor Relations (24-month horizon): review the macro narrative carried to investors, as a productivity revision would change the expected cost of capital.

The market still prices this transmission channel very little. The divergence between the Fed's independence rhetoric and the entry of private capital always resolves itself. The question remains how.

Any portfolio manager who ignores this shift is building portfolios on an obsolete neutral rate. The post-2020 rate regime closed the thirty-year bond bull run. Now AI introduces a second variable into the calculation of long-term real returns.

The Forecast

I offer a verifiable forecast, with an explicit horizon and indicator.

The Productivity and Jobs task force will publish at least one formal output — a report or recommendation — that will influence an official FOMC communication by December 31, 2027. Confidence: 65%. The falsification signal is clear: the absence of any publication or communication from the task force by that date.

This is a regime change in monetary governance, not an isolated episode.

What to Watch

Three indicators will confirm or refute the thesis over the coming quarters.

• Explicit references to AI productivity in FOMC minutes and Warsh's speeches.
• Adoption of analogous advisory bodies at the ECB, Bank of England, or Bank of Japan.
• Revisions to long-term productivity estimates in the Fed's economic projections.

The Bank for International Settlements is closely monitoring AI's impact on central banks, as covered by Central Banking on the BIS^[2] and Risk.net^[3]. The BIS has repeatedly flagged concentration risks in compute infrastructure.

Watch the background of the advisers, not just their reports. Governance reveals intent before documents do.