Year: 2026
Type: dissertation
Abstract: Banking organizations have increased their interest in integrating AI components into customer relationship management systems to improve efficiency and productivity since the revolution of AI technologies. These organizations operate in a highly regulated environments and there is limited understanding on how decisions are made on component level by responsible teams. Hence, this thesis explores how decisions are made on choosing how to handle new and existing AI components within CRM systems, whether to adopt, hybridize or replace, limited to Finnish banking environments. This thesis was conducted as a qualitative study, performing analysis on ten semi-structured interviews taken from individuals responsible for integrating AI components into CRM in Finnish banking organizations. Empirical analysis used thematic analysis with a deductive-inductive (Hybrid) coding approach. The empirical findings suggested that when decisions were made on choosing AI components, they would not be simple choices such as the best AI model or vendor, but rather a complex decision logic process to ensure compliance. Factors such as workflow and business value are primary objectives to determine an AI components worthiness to pursue. It is also further narrowed down by other elements such as data constraints, and governability constraints. When operating with AI components there are certain levels of uncertainty and risks that need to be verified or mitigated. Pilots, proof-of-concepts, reversibility, monitoring, fallback strategies, and human-AI accountability become crucial when AI moves from support to decision-making roles. In a broader spectrum, these findings may be transferable to similar regulated financial contexts. The main contribution of this thesis is a set of empirically derived decision heuristics for CRM integration teams who are making AI-enabled CRM component-level decisions in regulated banking environments. These heuristics translate the empirical concept governable configuration decisions, which captures how CRM teams prioritize AI components that fit business value creation and governance requirements that include compliance, maintenance, auditability, reversibility, accountability, over prioritizing feature richness alone. Theoretically, this thesis combines perspectives from CRM capability, socio-technical IS, AI governance, and decision making under uncertainty showing that value in regulated AI-enabled CRM depends on its governability. Keywords: Artificial intelligence (AI), Customer relationship management (CRM), Finnish banking, governable configuration decisions, decision heuristics, AI-enabled CRM, socio-technical factors, decision-making under uncertainty
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7223420331