Skip to content
Nerdy beta

From Technical Debt to Cognitive and Intent Debt

Engineering Technical Debt AI

Margaret-Anne Storey argues that generative AI is redistributing where the riskiest debt accumulates in a software system. Technical debt, the well-known cost of messy code, is no longer the only constraint, and AI is increasingly capable of helping pay it down. Two quieter forms have moved to the foreground: cognitive debt, the erosion of shared understanding inside the team; and intent debt, the missing goals, constraints, and rationale that both humans and AI agents need to know what the system is for. The triple-debt model treats software health as a property of code, people, and externalized intent in tandem.

Storey builds on Peter Naur’s “theory of the system,” Edwin Hutchins’s distributed cognition, and Shaw and Nave’s work on cognitive surrender, the way developers adopt AI output with minimal scrutiny while their confidence in the result inflates. This article anchors the Cognitive and Intent Debt concept page and supplies grounding for Costly Tech Debt and the Technical Investment workshop.

Resources