The Prompt-Kruger Effect: AI-Induced Overconfidence in Non-Programmer Software Builders
Year: 2026
Type: article
Abstract: This paper introduces and defines the term "Prompt-Kruger Effect": the confidence an AI coding agent gives a non-programmer that makes them believe they have become a software engineer. AI coding agents now let professionals from other fields, such as construction, finance, and operations, produce working software in hours. The visible output is mistaken for the invisible competence that software engineering requires: security, architecture, maintainability, error handling and debugging. The paper links the effect to the Dunning-Kruger effect and to recent empirical studies showing that AI assistance raises users' confidence faster than their actual quality of work, and proposes practical mitigations for individuals and organizations. The term "Prompt-Kruger Effect" was coined by Mohamed Ali El-Sayed and first defined in this paper.
Author: El-Sayed, Mohamed Ali
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225434572