AI Use Among Developers is Growing, but Trust Remains Low
According to Google Cloud’s DORA 2025 report, almost 90% of developers now use AI in their daily work—up from 76% in 2024. Developers use AI for everything from writing code and debugging to creating technical documentation.
But there’s a catch: only 24% of developers fully trust the results AI produces. On average, engineers spend about two hours a day using AI assistants—but most treat AI suggestions like “spam,” checking everything carefully. This mirrors a trend seen on Stack Overflow, where distrust of AI grew from 31% to 46% in just one year, despite its growing use.
Even at Google, AI is used cautiously. Ryan Salva, head of Gemini Code Assist, says developers rely on AI but always verify its work. CEO Sundar Pichai notes that these tools have boosted team productivity by 10%.
Dependence vs. Risk
AI has become essential: 65% of developers rely heavily on it, even if they don’t fully trust it. But there are risks. Security issues, like the “copypasta” attacks noted by HiddenLayer, show how AI tools can be misused.
To manage this, Google uses the DORA capability model, which sets clear rules and supervised workflows to prevent AI from running unchecked. The aim is simple: get the benefits of AI without creating new problems.
This debate isn’t just technical. Surveys show that people worry AI can reduce creativity and weaken social connections. Developers see AI as a brilliant but unpredictable “colleague.”
The Bottom Line
AI is now essential for software development, but distrust remains high. The challenge is building a future where AI is not just widely used, but also trusted. For now, AI is a powerful helper—but one that needs careful supervision.
