A recent informal survey of 103 developers shows AI coding assistants are mainstream tools that boost individual productivity but remain untrusted for production work and team collaboration.
Claude Code is the most preferred assistant (41%), followed by Cursor (21%) and GitHub Copilot (10%).
Developers treat AI as a productivity assistant for repetitive tasks, not an autonomous engineer for architectural or security-sensitive code.
From generating code snippets to debugging complex applications, AI-powered coding assistants are no longer experimental. Yet, despite this widespread adoption, one issue continues to define the relationship between developers and AI – trust. We recently ran an informal survey of 103 developers, and the results paint an interesting picture.
Instead of replacing developers, AI often removes repetitive or time-consuming tasks allowing engineers to focus on architecture, design decisions, and problem-solving. Because of this, developers continue to verify code with the same level of scrutiny they apply to code written by junior engineers. Looking Ahead The story of AI in software development is no longer about whether developers will adopt it – that question has already been answered.
They are everyday tools for millions of developers worldwide. At the same time, confidence in code remains surprisingly low, suggesting that while AI has become an essential assistant, it has not yet become a fully trusted teammate. Developers Are Using AI—Even If They Don't Fully Trust It Perhaps the most striking statistic is the gap between adoption and trust.
A year ago, the question every Enterprise Architecture (EA) leader asked about AI was s ome version of “What can it do?” Today, the question is, “Can I trust what it just did, and can I prove it to someone else?”
Originally published at nlocoding.com Only 18% of developers trust their AI debugging assistant to suggest production-ready fixes without review. The other 82%? They’re still glued to Stack Overflow, tabs multiplying like rabbits. (Source: JetBrains State of Developer Ecosystem 2026) Software eats itself faster every y
Originally published at nlocoding.com 81%of critical vulnerabilities exploited in 2025 had known fixes available for over 60 days. (CISA, 2026) Software doesn’t get hacked because attackers are clever. It gets hacked because maintainers move slow. That’s the ugly truth. AI is changing the speed equation—sometimes for b
Firms using AI firmwide nearly doubled this year, from 17% to 33%, and 90% of professionals grade their firm’s AI rollout an A or a B, according to Intapp’s 2026 Technology Perceptions Survey [1]. But 76% of professionals are using AI tools their firm never approved. Firms think they’ve solved this. Their own people ar
Security teams are starting to actively use AI for security work, including vulnerability triage, penetration testing, threat modeling, incident response, and code review. The promise is speed, but a security tool that moves fast and raises too many false alarms doesn’t save time. Engineers spend time on false alarms,
In this issue: Xinyue Chen, Mary Haasdyk Vooys, Henrike Lendowski, Glenda Sburelin, Adam Niklewicz, Melcher Oosterman, Ben Pearce, Susie Ang, Matt Rota, Anagh Banerjee.
Loading more related stories...
Open the app view to save this story, compare related coverage, and continue from the same source.
