A recent study from cloud software company Fastly has found that senior developers are much more likely to use AI-generated code compared with junior developers, as well as deploy this code for use in production.
Senior developers also feel that using AI tools makes them work faster, more than junior developers do. However, senior engineers were more likely to spend time correcting AI code compared with their less-experienced counterparts.
AI has without a doubt changed the landscape for developers, shaking up the job market and altering the day-to-day tasks of the role. While evidence is mixed on whether AI improves efficiency, this study certainly suggests that developers want to keep AI around for the time being.
Study Finds Senior Developers Use More AI-Generated Code Than Juniors
A new study from Fastly has found that senior developers are using more AI-generated code than junior developers, and are deploying more of this code to be used in production.
Overall, nearly one third of the senior developers polled revealed that over half of the code they ship is AI-generated. This figure is more than double the rate of junior developers, with only 13% reporting the same thing.
This just in! View
the top business tech deals for 2025 👨💻
Findings point to a higher rate of AI use among senior developers, and suggests there may be a higher level of trust in AI code for these experienced developers. Fastly’s study consisted of 791 professional developers, made up of senior developers (those with 10+ years of experience) and junior developers (those with less than two years’ experience).
Senior Developers Say AI Makes Them Faster, But Spend More Time Correcting Mistakes
AI is certainly good at taking over autonomous and repetitive tasks. However, the sectors and areas where it can save employees the most time is still not entirely clear. In the study, senior developers were more likely to view AI as a net time-saver.
In fact, senior developers had a stronger enthusiasm for AI as a tool to help them work faster, with 59% saying AI sped up their working process. Comparatively, only 49% of junior developers said the same.
However, these findings come hand-in-hand with the fact that senior developers were more likely to spend time correcting AI-generated code. This highlights a gap between how fast a developer believes they are working, and the reality of having to make changes.
Indeed, 28% of developers said they spent so much time editing AI code that any potential time-saving benefits were made void. Only 14% said they rarely needed to make big changes to work generated by AI.
Does AI Make Developers Faster and More Efficient?
Senior developers, the study points out, have plenty of experience to lean on when dealing with AI-generated code, compared with juniors. What may be the case is that senior developers are more capable of finding vulnerabilities or mistakes, and believe making these changes is more efficient than writing the code themselves from scratch.
However, it still may not be the case that tools such as Copilot or Claude make software engineers faster and more efficient. In another trial from Fastly, open-source developers took 19% longer to complete tasks when using code assistants. This suggests that what may feel fast to a developer — maybe because they are getting all of the work done, pretty much immediately — doesn’t equate to tangible productivity gains.
Another notable element of this study is its clear stance on job satisfaction. Roughly 80% of developers, across all experience levels, said coding felt more enjoyable when working with AI. Ultimately, while we may not have figured out yet how effective AI is for developers, they seem to want it to stay, for the time being.