Generative AI as a Driver of Organizational Innovation and Knowledge Creation

Generative AI (GenAI) tools produce ideas, drafts and analyses on demand, and organisations are adopting them to accelerate innovation. Whether they increase knowledge creation, the production of ideas that are novel, useful and non-redundant, is less clear: GenAI raises the quantity of ideas and lifts the novelty of less skilled users, but its outputs converge across users, which reduces the diversity on which organisational innovation depends. This paper models knowledge creation as the product of idea quantity, novelty, usefulness and non-redundancy, and compares three conditions in a simulation of 300 users with varying skill: unassisted ideation, GenAI-assisted ideation, and GenAI-assisted ideation with divergence prompts that direct the tool toward unconventional directions. GenAI assistance doubles idea quantity and raises novelty for low-skill users from 0.36 to 0.52, but convergence across users rises from 20% to 55%, so that net knowledge creation rises only from 1.33 to 1.50 and falls for high-skill users (2.12 to 1.97). Divergence prompting retains the quantity gain, raises novelty further and limits convergence to 35%, raising knowledge creation to 2.39 overall and benefiting both skill groups. The paper interprets the results through the knowledge-creation and creativity literatures, identifies convergence as the central risk of GenAI for innovation, and proposes practices for using GenAI to expand rather than homogenise an organisation’s idea space.

Hali tarjima qilinmagan