In AI-simulated Fed meeting, political pressure polarises board

A recent study simulating a Federal Reserve meeting with AI agents revealed that political pressure significantly polarizes board members and increases dissent, suggesting the Fed is only partially insulated from external scrutiny. While central banks globally are not yet using AI for monetary policy formulation, institutions like the Fed, ECB, and BoJ are increasingly adopting the technology to enhance operational efficiency, research, and economic analysis, albeit many remain in early adoption stages due to governance and data quality considerations.

Analysis

A recent study from George Washington University, which used AI agents modeled on Federal Reserve policymakers, indicates that political pressure significantly polarizes rate-setting deliberations and increases dissent. This simulation of a July 2025 FOMC meeting suggests that the Federal Reserve's operational independence is only partial and that external scrutiny can materially shape internal decision-making. While no central bank is currently using AI to formulate monetary policy, the technology is being increasingly adopted for analytical and efficiency purposes. The Fed is researching its use for analyzing minutes, the ECB employs machine learning for inflation forecasting, and the Bank of Japan has used large language models to identify a potential shift in inflation drivers from raw material to labor costs. However, a Bank for International Settlements report notes that most central banks are still in the "initial adoption phase," with progress tempered by the need for adequate governance and high-quality data.

Key Decisions for Investors

• Investors should monitor political pressure on the Federal Reserve, as the simulation suggests it can lead to increased dissent among policymakers and potentially less predictable monetary policy outcomes.
• The use of AI by central banks for deeper economic analysis, such as the BoJ's identification of shifting inflation drivers, implies that policy decisions may become more responsive to nuanced, forward-looking data sets, requiring investors to adjust their own analytical focus.
• Recognize that the integration of AI into central banking is a gradual, long-term trend, not an immediate market disruptor, given that significant hurdles in governance and data quality remain.