AI changes who owns code and who is responsible for mistakes
It Is Not My Code Anymore
Software EngineeringArtificial Intelligence
Summary
When AI helps write computer programs, it can be confusing to decide who actually made the code and who should fix problems if something goes wrong. The authors explain that just knowing who wrote a piece of code isn’t enough to figure out who should review it or decide when it’s ready to release. They suggest focusing on whether the code meets the goals it’s supposed to achieve, and considering the whole process that made it, not just the final product. This approach also helps compare AI systems that create code to ones that perform tasks directly without writing code.
What this means in practice
- •For software quality engineers: Create evaluation criteria focusing on service outcomes rather than just code authorship when testing AI-generated code.
- •For software release managers: Define release responsibilities based on collective ownership and process checks, not solely on who generated code segments.
A position paper. It proposes an approach and reports no results.
Authors
Augusto Camargo
Abstract
AI-assisted programming raises distinct questions about who produces code, who feels ownership of it, and who is responsible when it fails. This research note examines these distinctions through a hypothetical enrollment failure and a selective reading of the literature. Identifying the producer of a defective expression does not, by itself, determine the duties of reviewers, release decision-makers, or service operators. Collective ownership likewise leaves those duties to be specified. The discussion then considers how quality engineering can evaluate both generated implementations and the processes that produce them. Acceptance criteria should be justified by the required service outcome, with component checks contributing evidence toward that outcome. This perspective also permits comparison with systems that perform a task without generating a separate program for it. Such substitution would change the object of authorship while leaving the service obligation intact. The note reports no new empirical results; it proposes distinctions and evaluation questions for AI-assisted software production.