Paper 04 · The gap is named

AI Optimizes. The Bank Pays the Price. Decision Integrity in Agentic Retail Banking

The paper in which the Fiduciary Gap™ is named and defined, and Decision Engineering™ is named as the discipline.

Deepak Aggarwal · Posted 17 February 2026 · 20 pages
SSRN 6199979 · DOI 10.2139/ssrn.6199979

What this paper argues

Retail banking has been at the forefront of AI adoption. Optimisation systems now hold real executional authority over pricing, retention and deposit behaviour — authority granted faster than the institution's obligations were encoded into them.

This paper names and defines the Fiduciary Gap™: the mismatch between what the AI system is optimised to do and the broader responsibilities of the institution towards customers and balance sheet. It also names the discipline — governance that is no longer concerned with validating models, but with engineering decision authority.

The deposit-optimisation case at the centre of the paper shows the pattern in full: a win on the dashboard, a loss on the balance sheet.

Terms used here are defined once in the glossary: The Fiduciary Gap™ · Decision Engineering™

Start with one decision

Name an important decision made in the last ninety days that your institution could not fully replay today — the rule that applied, the data that was live, the authority that permitted it.