Bridging the governance gap: adapting technology pioneer ai governance frameworks for financial services institutions

Financial services institutions are deploying sophisticated GenAI and agentic AI systems within governance frameworks designed for traditional statistical models, creating a widening gap between AI risk and institutional capacity to govern it. This thesis addresses that gap by translating lessons from AI technology pioneers into the financial services context. Using a design science research methodology, the first study examines governance documentation from five technology pioneers: IBM, Microsoft, Google DeepMind, OpenAI, and Anthropic. An eight-dimensional analytical framework reveals 95% convergence on four foundational practices: risk-based classification, lifecycle governance, adversarial testing, and deployment authority with binding decision rights. A structured gap analysis then uses these pioneer frameworks as a baseline to identify systemic weaknesses in financial services governance, including near-zero adversarial testing, AI decisions made at the business-unit level without governance oversight, and reliance on periodic review mechanisms inadequate for managing AI behavioral drift. Informed by these findings, the thesis proposes the KARMA (Knowledge, Agency, Responsibility, Machine Accountability) framework, which merges pioneer-derived mechanisms with financial services regulatory requirements and the three lines of defense model. KARMA specifies governance structures, decision authorities, risk tiering criteria, lifecycle processes, and control standards for implementation within existing institutional architectures. A simulated banking case study illustrates its operational application. Validation through a single-round Delphi assessment with eight domain experts produced unanimous consensus across all 21 evaluation items, with all panelists recommending pilot testing. The thesis contributes the most comprehensive comparative analysis of technology-pioneer AI governance to date, establishes the first explicit governance baseline derived from those practices, and conducts the first systematic gap analysis of financial services AI governance against it. Expert practitioners endorse KARMA as conceptually sound, operationally complete, and ready for implementation, enabling financial services organizations to build more robust approaches to AI governance.