Professional Software Developers Don't Vibe, They Control: AI Agent Use for Coding in 2025

2025-12-15

AI Summary

This study investigates how professional software developers in 2025 deploy AI agents in real-world engineering contexts, focusing on their roles, human–agent collaboration strategies, and the delineation of human–machine boundaries. Method: We employed a mixed-methods approach—field observations (N=13), in-depth interviews, and open-ended surveys (N=99)—grounded in software engineering practice frameworks and human–agent collaboration theory. Contribution/Results: We introduce the “Developer-Primary Paradigm,” empirically demonstrating that developers retain quality-driven design authority and strategic decision-making, treating AI agents as controllable productivity enhancers—not autonomous replacements. We identify six high-fit agent tasks (e.g., boilerplate generation, documentation completion) and four critical human-only responsibilities (e.g., architectural design, root-cause analysis). Based on these findings, we propose actionable agent governance strategies, principled accountability allocation, and interface design guidelines—challenging prevailing narratives of fully autonomous agents and providing an evidence-based foundation for AI-augmented software engineering.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Multiagent Systems: Agent/AI Theories and Architectures
• Humans and AI: Planning and Decision Support for Human-Machine Teams

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search

Abstract

The rise of AI agents is transforming how software can be built. The promise of agents is that developers might write code quicker, delegate multiple tasks to different agents, and even write a full piece of software purely out of natural language. In reality, what roles agents play in professional software development remains in question. This paper investigates how experienced developers use agents in building software, including their motivations, strategies, task suitability, and sentiments. Through field observations (N=13) and qualitative surveys (N=99), we find that while experienced developers value agents as a productivity boost, they retain their agency in software design and implementation out of insistence on fundamental software quality attributes, employing strategies for controlling agent behavior leveraging their expertise. In addition, experienced developers feel overall positive about incorporating agents into software development given their confidence in complementing the agents' limitations. Our results shed light on the value of software development best practices in effective use of agents, suggest the kinds of tasks for which agents may be suitable, and point towards future opportunities for better agentic interfaces and agentic use guidelines.

Problem

• Investigates experienced developers' use of AI agents in software development
• Examines motivations, strategies, and task suitability for AI agent integration
• Explores how developers maintain control and ensure software quality with agents

Innovation

• Experienced developers control AI agents using expertise.
• They retain agency for software quality and design.
• Agents are used for productivity with controlled tasks.