A recent study by academics at George Washington University simulated a Federal Reserve meeting using artificial intelligence (AI) agents. These AI agents were modeled on real-life policymakers, with their behaviors based on the historical policy stances, biographies, and public speeches of each Federal Open Market Committee (FOMC) member. The AI-driven FOMC was then tasked with making decisions based on real-time economic data and financial news.
The findings showed that when political pressure was introduced, the AI agents became more divided, and dissent became more common. The researchers concluded that this simulation demonstrates that the Federal Reserve is not entirely insulated from political influence. External scrutiny, they noted, can shape internal decision-making, even within an institution guided by formal rules and protocols.
