Stack Overflow Launches Survey on AI Agent Usage and Human Autonomy
Stack Overflow has launched a new survey to investigate how software engineers are integrating AI agents into their daily workflows. Following previous data showing that 31% of developers used AI agents in 2025, the platform seeks to understand the evolving dynamic between human developers and autonomous tools. The survey explores whether professionals remain "human-in-the-loop," actively writing code, or have shifted to "human-on-the-loop," primarily overseeing and editing agent-generated output. Key questions address the level of autonomy granted to agents, pricing models such as pay-per-token versus flat rates, and the overall effectiveness of these tools in improving software quality, including testing outcomes. This initiative aims to define new boundaries of responsibility in software development as AI capabilities expand from simple auto-completion to building entire tools. The survey is open until April 30, 2026, with results scheduled for publication in May. By gathering this data, Stack Overflow intends to provide insights into how the role of the software engineer is transforming in an era of increasingly capable artificial intelligence, helping the community adapt to these technological shifts.
Wire timeline
Stack Overflow Launches Survey on AI Agent Usage and Human Autonomy
Stack Overflow has launched a new survey to investigate how software engineers are integrating AI agents into their daily workflows. Following previous data showing that 31% of developers used AI agents in 2025, the platform seeks to understand the evolving dynamic between human developers and autonomous tools. The survey explores whether professionals remain "human-in-the-loop," actively writing code, or have shifted to "human-on-the-loop," primarily overseeing and editing agent-generated output. Key questions address the level of autonomy granted to agents, pricing models such as pay-per-token versus flat rates, and the overall effectiveness of these tools in improving software quality, including testing outcomes. This initiative aims to define new boundaries of responsibility in software development as AI capabilities expand from simple auto-completion to building entire tools. The survey is open until April 30, 2026, with results scheduled for publication in May. By gathering this data, Stack Overflow intends to provide insights into how the role of the software engineer is transforming in an era of increasingly capable artificial intelligence, helping the community adapt to these technological shifts.
Stack Overflow Blog