AI adoption in finance: the latest developments

Finance is the most mature AI adopter, and the gap between leaders and laggards is widening. The biggest development in 2024-2025 is the use of LLMs for regulatory reporting and compliance. JPMorgan and Goldman have both deployed internal AI systems that draft regulatory submissions and compliance documentation. The time savings are 50-70% on the drafting work, but the review process is still critical. The risk is real: an AI-drafted regulatory filing with an error would be catastrophic. The organizations that are succeeding are the ones that treat AI as a drafting assistant, not a replacement for compliance officers.

The retail banking side is seeing the most visible consumer-facing AI. Chatbots have moved from the FAQ tier to genuine account management assistants. The best implementations can handle balance inquiries, transaction disputes, and even basic financial advice. The challenge is the trust gap: customers are still hesitant to share sensitive information with an AI. The banks that are winning are the ones that offer a seamless handoff to a human when the conversation crosses a complexity threshold. The AI handles the routine, the human handles the nuance.

The fintech side is where the most interesting AI applications are happening. Algorithmic trading has used AI for years, but the new frontier is using LLMs for investment research and sentiment analysis. The challenge is that LLMs are good at summarizing information but not at predicting market movements. The startups that are winning are the ones that use AI for information processing and keep the investment decisions human-driven. The AI makes the analyst faster, not the analyst obsolete.