AI-generated code accounted for more than half of shipped code among nearly one-third of senior developers who participated in a recent survey by cloud platform provider Fastly.
The company’s July 2025 survey of 791 professional developers found that 32% of senior developers (those with 10 or more years of experience) said more than half of their shipped code was AI-generated. Among junior developers, just 13% said the same. Fastly’s findings suggest that more experienced engineers not only are using AI more aggressively but also may trust it more in production environments. This is surprising given growing concerns about vibe coding introducing vulnerabilities into code, said Fastly.
Nearly one in three developers (28%) said they fixed or edited AI-generated code frequently enough to offset most of the time savings. Only 14% said they rarely needed to make changes. And yet, more than half of developers said they felt they worked faster with AI tools like Copilot, Gemini, or Claude. “AI will bench-test code and find errors much faster than a human, repairing them seamlessly. This has been the case many times,” one senior developer wrote. A junior respondent cited trade-offs: “It’s always hard when AI assumes what I’m doing and that’s not the case, so I have to go back and redo it myself.”
