fotograzia/Getty Images
Programmers are using AI more than ever, but they don’t like or trust the tools very much, according to the 2025 Stack Overflow Developer Survey.
The survey of almost 50,000 developers found that 84% now use or plan to use AI tools in their workflow, up from last year (76%). Over half of professional developers (51%) use these tools daily.
Also: The best AI for coding in 2025 (and what not to use)
Such figures might suggest that programmers must love AI. However, only 60% expressed positive sentiment toward AI tools, a proportion down from over 70% in both 2023 and 2024.
Distrust is a defining theme of the survey. In 2024, 43% of developers felt good about AI accuracy, and only 31% were skeptical. By 2025, 33% of developers trusted AI tool outputs, 46% expressed active distrust, and a mere 3% said they highly trusted the results. Among seasoned professionals, the “highly trust” figure dropped to just 2.6%, with 20% reporting strong skepticism.
Also: Bad vibes: How an AI agent coded its way to disaster
In short, developers certainly use AI, but trusting the technology to get the job right on its own is another matter. And after IT leader Jason Lemkin’s experience with a Vibe programming project that disintegrated, taking with it a production database, who can blame them?
The Stack Overflow study also found that the biggest single frustration, cited by 66% of developers, is dealing with “AI solutions that are almost right, but not quite,” which often leads to the second-biggest frustration, at 45%, being “Debugging AI-generated code is more time-consuming.”
As Bill Harding, CEO of Amplenote and GitClear, noted in GitClear’s AI Copilot Code Quality study, which analyzed 211 million lines of code, “developers trust the current generation of AI assistants about as much as we trusted the previous generation, i.e., not much.”
It’s not just programmers who don’t trust AI. A recent survey of more than 1,100 Americans revealed that only 8.5% said they “always trust” the information they receive from Google’s AI Overviews, and 21% said they have zero trust in the feature’s ability to surface accurate information. A new KPMG study found that, worldwide, 66% of people use AI, but only 46% trust AI systems.
Also: How to get rid of AI Overviews in Google Search: 4 easy ways
Junior developers who trust AI the most could be making a big mistake. A popular blog post by Namanyay Goel, an independent developer, warned: “We’re trading deep understanding for quick fixes, and while it feels great in the moment, we’re going to pay for this later.”
Worries over AI-created technical debt are growing. Harding warned that if companies continue to measure developer productivity by the simple-minded metric of number of commits or lines of code written, AI-driven technical debt will spiral out of control. “Leaders need to recognize that more code is often worse,” he said, suggesting that copying and pasting code leads to higher defect rates. Indeed, GitClear found a direct connection between the rising defect rate and AI adoption.
All that said, Stack Overflow also revealed in its survey that OpenAI’s GPT models are the most popular large language models, with 82% of developers who use AI indicating that they used them for development work in the past year. Anthropic’s Claude Sonnet models came second, followed by Google’s Gemini Flash.
Despite AI’s rise, when it comes to integrated development environments (IDEs), programmers still prefer Visual Studio (75%) and Visual Studio Code (29%) over AI-first programming IDEs. The old-school, simple code editors, Vim and Notepad++, remain popular, even among programmers who use AI. That said, Microsoft’s incorporation of Copilot into its tools has proven to be a smart move.
Also: Coding with AI? My top 5 tips for vetting its output – and staying out of trouble
As before, JavaScript, HTML/CSS, and Python maintain their status as the most widely used programming languages. Python, and probably due to the popularity of Python-based generative AI libraries TensorFlow and PyTorch, is especially sought after by developers adopting a new language. However, Rust, with an 83% approval mark, remains the most admired language.
Looking ahead, while AI is being adopted quickly, developers are, if anything, growing cautious about handing off critical tasks to agents. A vast majority (75%) said human advice is still irreplaceable in scenarios where they don’t trust AI’s output.
As for AI agents, they have yet to reach mainstream acceptance. Over half of the survey’s respondents use simpler AI tools, and 38% have no plans to adopt agents soon.
