Harnessing Generative AI for Economic Insights A recent study by researchers from Georgia State University and Chicago Booth demonstrates how generative AI can extract valuable economic forecasts from corporate conference calls. By analyzing over 120,000 transcripts, the team created the AI Economy Score, which robustly predicts future GDP growth, production, employment, and wages—outperforming traditional forecasting methods. The Challenge: Corporate managers have significant economic impact. While their expectations contain valuable economic insights, traditional surveys are costly and limited in scale. The Solution: Using ChatGPT to analyze earnings call transcripts, researchers developed a methodology to extract managerial expectations about economic conditions, creating a panel of forecasts from 5,513 unique companies to create the AI Economy Score. When comparing predictive power, the AI Economy Score demonstrated 32.5% higher in-sample fit than SPF forecasts (one of the oldest and most respected quarterly surveys of economic forecasts in the United States) for GDP growth. This superior performance suggests that managerial expectations captured through AI analysis contain unique information not available in traditional forecasting methods. The AI Economy Score provides a complementary tool to existing economic forecasting approaches. As AI continues to evolve, we can anticipate further advancements in economic analysis and decision-making tools. The AI Economy Score represents a significant step forward in harnessing the power of generative AI for economic insights, providing valuable information for both macroeconomic and microeconomic decision-making.