
A recent Bank Director survey reveals that the majority of banks are opting to rent agentic AI solutions through third-party vendors rather than developing them in-house. While 72% of banks have adopted generative AI, only 30% utilize autonomous agents, with many relying on external software to manage reasoning and data access. This trend raises concerns regarding auditability and control in the highly regulated financial sector. In response, developers Francesco Galletta and Felippe Felisola Caso have taken a different approach by building a custom, trust-focused financial analyst harness. Their work, featured at a recent workshop, highlights the limitations of off-the-shelf agents and the growing movement toward bespoke, internal AI development. Other industry experts, such as Ignacio Martinez, are also advocating for custom-built solutions, specifically for complex tasks like financial crime detection, emphasizing the importance of memory-aware agent systems in maintaining security and transparency within banking operations.
This is a summary. Read the full article at the original source:
Dev.toRelated stories
This article examines a critical issue in a payment service: despite passing code reviews and having successful tests, users experienced double charge…
Review of Igor Shimko's book 'I Will Teach You to Manage Money': Relevance for IT Professionals
Amid discussions regarding pension strategies and the instability of global systems, the author reviews Igor Shimko's book, 'I Will Teach You to Manag…
Online retailers that don't support digital wallets are losing Gen Z customers
A recent study highlights a growing trend among younger consumers, specifically Gen Z, who are increasingly prioritizing convenience in their online s…



