We develop a novel measure of trust in the Federal Reserve using Generative Artificial Intelligence to analyse millions of tweets about the Fed, its leadership and its policy framework and decisions. Our measure reacts in an intuitive way to various macro-financial variables and indicators of U.S. monetary policy. To study the effects of trust shocks, we use a narrative identification approach based on ethical scandals embroiling some FOMC members, and we study the effects of these shocks using a daily VAR. We find that trust shocks have highly persistent effects on macroeconomic variables despite having short-lived effects on our trust measure: they weaken business conditions, the stock market and news sentiment, while increasing the VIX index. Inflation expectations also increase following a trust shock, worsening the inflation-output trade-off.