Agentic AI Is Reshaping How Banks Handle Risk
Autonomous AI systems are now analyzing credit risk, detecting fraud, and making investment decisions. Wall Street is watching.
The Financial Times ran a deep dive this week on 'Agentic AI and the Shift to Autonomous Finance' — and the implications are staggering. Banks are deploying AI systems that don't just assist human decision-making; they make decisions autonomously, in real-time, at scales no human team could match.
What Is Agentic AI in Finance?
Unlike traditional AI that flags potential fraud or suggests credit limits, agentic AI takes actions. It can: approve or deny loans within milliseconds, adjust portfolio allocations based on market signals, issue fraud alerts that automatically freeze accounts, and renegotiate terms with counterparties.
Goldman Sachs, JPMorgan, and Citadel have all deployed versions of this technology. The efficiency gains are enormous — but so are the risks.
The Risk Nobody Is Talking About
When an autonomous AI system makes a bad decision, it doesn't make it once. It makes it thousands of times per second, across every market, before a human can intervene. The flash crash of 2010 was caused by a single bad algorithm. Agentic AI multiplies that risk by orders of magnitude.
Wall Street has always been about speed and scale. Agentic AI delivers both — along with correlated failures that could take down entire markets in minutes.
The Regulators Are Behind
The SEC and FCA have issued discussion papers on AI governance. Actual rules remain sparse. In the meantime, banks are essentially self-regulating their most consequential automated systems. History suggests that ends poorly.