What is it about?
As artificial intelligence and automation transition from simple tools to active participants in software development by handling tasks like code generation, testing, vulnerability repair, and deployment, new challenges arise. While ethical and trustworthiness principles such as transparency, fairness, and accountability are widely discussed, turning these high level principles into actual executable engineering behavior has remained a missing link. In this paper, we introduce Software Engineering Governance as a vital organizing perspective. Rather than viewing governance merely as an administrative task, we position it as a continuous cycle structured around automated execution. Before execution, trustworthiness objectives are translated into operational lifecycle artefacts that guide or constrain engineering actions. After execution, the resulting operational evidence is evaluated to ensure alignment with original goals and to support organizational accountability. By connecting principles, execution, and evaluation, this work provides a foundational framework to help bridge the gap between abstract artificial intelligence ethics and reliable trustworthy automated software engineering.
Featured Image
Photo by Mohamed Nohassi on Unsplash
Why is it important?
This work addresses a timely challenge as automated systems and software agents rapidly shift from simple assistants to active decision makers across the software lifecycle. Because traditional organizational governance struggles to keep pace with autonomous engineering behavior, this paper provides a theoretical and foundational perspective. Instead of offering a closed implementation framework, it introduces a conceptual structure that connects high level trustworthiness principles to automated execution and evidence evaluation. By organizing these complementary responsibilities into a continuous cycle, the work maps out the core relationships between operationalization and verification while deliberately leaving open key methodological challenges for future research in automated software engineering.
Perspectives
This paper started from a very practical realization as we watched artificial intelligence tools rapidly evolve from simple coding assistants into autonomous engineering actors across the software lifecycle. While discussing this shift, we often joked about managing a chaotic mess of agents in modern development environments, which highlighted a glaring gap. Everyone talks about high level ethical principles, but nobody has a clear way to govern what these automated agents are actually doing. Writing this paper was our attempt to step back, stop treating governance as a purely administrative checklist, and start building a foundational bridge between abstract trustworthiness objectives and real software execution. We hope this conceptual starting point sparks a broader conversation in the community and helps researchers tackle the open challenges of trustworthy automated software engineering.
Carlos Mario Braga Ortuño
Universidad de Castilla-La Mancha
Read the Original
This page is a summary of: Software Engineering Governance: Foundations for Trustworthy Automated Software Engineering, October 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3820756.3844918.
You can read the full text:
Resources
Contributors
The following have contributed to this page
