Over the past several years, the productivity gains of AI have been touted left and right, but just because AI can generate code doesn’t necessarily mean that it helps speed up the software development life cycle.
According to a report from GitLab, an “AI Paradox” has emerged. “While AI accelerates coding, fragmented toolchains and new compliance complexities are creating bottlenecks that cost teams nearly a full workday per team member each week,” the company wrote.
GitLab’s research, which gathered responses from over 3,000 DevSecOps professionals, found that those workers are losing 7 hours per week to inefficient processes, such as a lack of cross-functional communication, limited knowledge sharing, and use of different tools across teams.
Additionally, 60% of respondents use more than five tools for software development and 49% use more than five AI tools.
GitLab believes the solution to these issues lies in following platform engineering approaches to address requirements for AI orchestration, governance, and compliance. Eighty-five percent of respondents believe agentic AI will be successful if it’s implemented in this way.
The report also revealed that AI will create more engineers, rather than replacing existing ones. Seventy-six percent of respondents believe that as AI coding gets easier, there will be more engineering roles.
A majority of respondents also see the increased pressure to upskill as their roles evolve, with 87% wishing their companies invested more into helping them do so.
Other findings of the report are that only 37% of respondents would trust AI to handle tasks without human review, 88% agree there are essential human qualities that agentic AI can’t fully replace, and 70% say that AI is making compliance management more challenging.
