ASE-26: Establishing Agentic Software Engineering as an Academic Discipline

The nature of professional software engineering is undergoing a fundamental transformation, with practitioners increasingly focusing on directing AI agents rather than manually writing code. This transition is no longer speculative; empirical data spanning several years confirms its reality. According to Anthropic’s Economic Index, automation now accounts for 79 percent of interactions involving Claude Code [2]. Furthermore, research by Handa and colleagues at Anthropic indicates that AI exposure encompasses approximately 75 percent of the distinct activities within the Computer Programmer role [3]. In the labor market, this shift manifests in a 13 percent relative decline in employment among workers aged 22 to 25 in occupations most susceptible to AI, as reported by Brynjolfsson and colleagues at Stanford’s Digital Economy Lab [4].

However, this evolution remains incomplete. Current academic discourse on agentic software engineering suggests that the primary deficit is not a lack of advanced AI models, but rather the absence of structured practitioner discipline. To address this gap, this paper introduces ASE-26, a comprehensive undergraduate curriculum designed to formalize agentic software engineering as a distinct discipline. The curriculum is archived as a citable reference on Zenodo under the CC BY-ND 4.0 license [12].

The paper outlines the disciplinary framework underpinning the curriculum and details its conceptual contributions, most notably the "evolutionary spiral," which serves as the operational model for the co-evolution of intent and construction. It further describes the twenty-one-module structure used to organize the discipline for educational purposes. Additionally, the text explores the pedagogical implications of grading work co-created with agents, the competencies graduates will possess upon completion, and the design strategies ensuring the discipline remains relevant beyond the specific capabilities of current models.

The central argument posits that the industry’s current skills shortage aligns precisely with the competencies defined by this discipline. Consequently, the paper asserts that structured undergraduate curricula in agentic software engineering represent the primary mechanism for bridging this professional gap.

Source: arXiv Generated at: 2026-06-02 00:00:00 UTC