AI Agents Could Spark Bank Runs, Economist Warns

An economist at Apollo Global Management has issued a warning that agentic artificial intelligence could potentially trigger bank runs. The concern, articulated in a brief note, centers on the ability of advanced AI agents to rapidly and simultaneously withdraw funds from financial institutions. This hypothetical scenario posits that if a large number of individuals or entities were to deploy agentic AI to manage their finances, and if these agents were programmed or instructed to move funds en masse in response to perceived economic instability or a specific trigger, it could lead to a sudden and overwhelming demand for withdrawals. Such a coordinated and rapid outflow of capital could destabilize banks, even those that are fundamentally sound, by depleting their liquid reserves. The economist's perspective highlights a novel risk associated with the increasing sophistication and autonomy of AI systems, extending beyond traditional concerns about job displacement or data privacy to systemic financial stability. Agentic AI refers to AI systems that can autonomously perceive their environment, make decisions, and take actions to achieve specific goals without continuous human intervention. These agents could be programmed to monitor market conditions, news, or specific financial indicators and execute transactions, including fund transfers, based on predefined criteria or learned behaviors. The speed at which these agents can operate, coupled with their potential to coordinate actions across numerous accounts, presents a unique challenge for traditional banking systems, which rely on predictable customer behavior and manageable withdrawal rates. The implication is that the speed and scale of AI-driven financial actions could outpace the capacity of banks to manage liquidity, potentially leading to a crisis. This warning from a prominent economist at a major private equity firm underscores the growing need for regulators and financial institutions to consider the broader economic and societal impacts of advanced AI technologies. It suggests that future financial regulations may need to account for the unique risks posed by autonomous AI agents, possibly by introducing new protocols for managing digital asset flows or by developing mechanisms to monitor and control AI-driven financial activities. The scenario, while hypothetical, serves as a call to action for proactive risk assessment and mitigation strategies in the face of rapidly evolving AI capabilities. The economist's viewpoint suggests that the interconnectedness facilitated by AI could amplify financial shocks, making the traditional banking system more vulnerable to rapid, large-scale digital runs than previously understood. This perspective is particularly significant coming from an institution like Apollo Global Management, which is deeply involved in financial markets and asset management, indicating that these concerns are being considered at the highest levels of the financial industry.