AI Where It Matters: Where, Why, and How Developers Want AI Support in Daily Work

AI adoption varies dramatically by task type: Developers show strong current use and demand for AI improvement in core technical work like coding and testing, but explicitly want to limit AI involvement in mentoring and other relationship-centric activities, suggesting one-size-fits-all AI deployment is inappropriate for software development.

Different tasks require different responsible AI safeguards: The study identifies context-specific priorities—reliability and security matter most for systems-facing tasks, while transparency and user control are critical for general coding work, and fairness/inclusiveness become paramount for human-facing applications, indicating that responsible AI governance must be tailored rather than universal.

Developers see AI as a tool to reduce tedious work: There is high demand for AI to handle documentation, operations, and other "toil" tasks, revealing that responsible AI implementation should focus on augmenting human work by eliminating drudgery rather than replacing core decision-making.

Maintaining human control and understanding is a priority: Developers emphasize the need for transparency, alignment, and steerability—wanting to understand and direct AI outputs rather than accept them blindly, which has implications for how AI tools should be designed to keep humans in the loop.

This is the first large-scale empirical evidence of developer preferences: With 860 developers surveyed using mixed methods, this research provides concrete, data-driven guidance for AI tool designers and organizations, moving beyond speculation about where AI should be deployed in software development.