The Buy-or-Build Decision, Revisited: How Agentic AI Changes the Economics of Enterprise Software

David Klotz's arXiv paper (2604.26482) systematically re-evaluates the enterprise make-or-buy decision in the age of agentic AI—systems capable of autonomous code generation. Combining transaction cost economics and the resource-based view, Klotz analyzes seven canonical decision factors: cost, strategic differentiation, asset specificity, vendor lock-in, time-to-market, quality/compliance, and organizational capability. He argues that AI shifts the Make option from a pure hierarchy to a hybrid governance form that blends code ownership with external AI infrastructure dependency, fundamentally altering its economics and risk profile.

Klotz develops a typology of enterprise applications based on their sensitivity to AI-driven shifts. Commodity utilities and custom differentiating apps emerge as the strongest candidates for in-house building, as AI dramatically cuts development time and cost. However, regulated and mission-critical systems—where compliance, reliability, and auditability are paramount—remain firmly in the buy domain. The 'SaaSocalypse' narrative, which predicts AI will render most SaaS obsolete, is overstated: for the majority of enterprise categories, buying software still wins on risk-adjusted total cost of ownership.