Dr. Kollin Napier presented at the Mississippi Bankers Association’s 2026 Wealth Management Conference on October 28 in Ridgeland, discussing how AI is being used in banking and finance and what teams can do with it in their own work.
Napier, chief artificial intelligence officer for the Mississippi Artificial Intelligence Network (MAIN), Mississippi’s statewide AI initiative, spoke at MBA headquarters. The presentation, “The State of AI: Banking, Finance, and Mississippi’s AI Leadership,” connected examples from financial services with the tools, training, and guidance available through MAIN.
AI around the client meeting
The work before and after a client meeting provided a way to look at AI’s role in finance. Napier included Meeting Journey, which Merrill and Bank of America Private Bank announced in March. Bank of America describes the system as helping advisors prepare for meetings, summarize virtual meetings with client consent, and organize decisions and follow-up tasks.
The presentation also considered clients who bring an answer they found online or through an AI tool to their advisor. The advisor still needs to understand the client’s circumstances and help make sense of that information.
A brief fictional disclosure comparison brought the discussion back to the documents themselves. One excerpt stated an advisory fee but said nothing about custody charges. The exercise showed why missing information needs to remain an unanswered question instead of becoming a claim that a charge does not apply. The example came from a task in MAIN’s finance prompting guide; its figures and document excerpts were created for training.
When AI can take the next step
Napier also discussed AI agents, which can choose steps and use connected tools to carry out a task. In the presentation’s research example, an agent could gather approved information and prepare a draft. Permission to send a client message or change a record would be a separate decision.
That discussion included the information a tool can retain in prompts, answers, logs, and provider records. The presentation addressed approving the particular tool and setup for the intended use, with risk and compliance teams involved in decisions about client information.
AI-assisted fraud was another concern. The session considered requests that sound familiar but still need verification through an established contact route, especially when someone is pressing for urgent action.
Resources for banking teams
MAIN’s resources gave the discussion practical starting points. The finance policy template helps organizations work through data rules and approval responsibilities. It requires adaptation to the institution and does not certify compliance. The AI Agents Playbook provides guidance for planning a small pilot with approved inputs, clear access limits, and someone responsible for reviewing the work.
The proposed pilot returned to the document-comparison task. Its evaluation included useful results and errors, along with the time spent checking and correcting the draft. The presentation’s timing example was illustrative, not a report of measured savings.
Napier also shared MAIN’s free, online, self-paced AI training for Mississippi participants, including AI in Business. Seats and licenses are limited, and credentials have separate requirements. The training and guides are available to teams exploring how AI might fit their work.
MAIN thanks MBA Programs Manager Lori Richardson and the Mississippi Bankers Association for coordinating the session and inviting Napier to participate.