In AI-simulated Fed meeting, political pressure polarises board

A simulated Federal Reserve meeting that used artificial intelligence agents modeled on real-life policymakers showed political pressure polarised members of the board in their rate-setting deliberations.

In the study released on August 31, academics at George Washington University simulated a Federal Open Market Committee meeting using AI agents modeled by each member based on their historical policy stances, biographies and speeches.

The AI FOMC was then made to process real-time economic data and financial news to reach a decision.

The findings…