LLMs Playing Fed Officials Vote to Raise Rates

In one line: The Financial Times cast LLMs as members of the Federal Reserve's FOMC and had them decide on interest rates — and the models reportedly voted to raise.

Key points

• The FT explored a simulation in which LLMs stood in for members of the Fed's Federal Open Market Committee (FOMC).
• Given economic indicators and asked to vote like policymakers, the models reportedly leaned toward a rate hike.
• It is an example of using AI not for actual policy, but as a tool to observe and compare macroeconomic judgment.

Why it matters

The experiment offers a glimpse into how LLMs reason toward policy decisions from data, and how that differs from human experts. As AI is increasingly floated as an aid for financial and economic analysis, it is a useful lens for examining the models' biases and limits.

More

• The LLM FOMC votes to raise — Financial Times