Misguided Corporate AI Promotion: How Shareholder-Oriented Storytelling Can Distort Technology Selection

Abstract

This paper examines how corporate AI promotion can become detached from rational technology selection. Generative AI is valuable for tasks involving contextual understanding, translation, summarization, text generation, and ideation. However, many deterministic business processes may be better served by conventional programs, RPA, macros, or dedicated systems. The paper argues that companies should ask not only whether AI can be used, but whether AI is actually necessary. In particular, it proposes examining whether a task truly requires the probabilistic and context-dependent inference characteristic of generative AI. The paper also explores how shareholder-oriented storytelling, internal AI-related KPIs, and the pressure to demonstrate AI initiatives may contribute to making AI utilization an objective in itself. The central argument is that corporate AI promotion should be evaluated by the business value created and the rationality of technology selection, rather than by the volume of AI adoption. Not using AI for tasks where AI is not the better choice is also a legitimate AI strategy.