Mike Vizard summarizes findings from a recent GitLab DevSecOps survey, revealing how AI coding is driving demand for software engineers and transforming DevOps, with significant implications for compliance and team collaboration.

How AI Coding Is Shaping Software Engineering and DevOps Roles

Overview

A recent GitLab survey of 3,266 DevSecOps professionals—conducted by Harris Poll—explores how the adoption of AI tools is transforming software development and DevOps practices.

Key Findings

AI Tool Adoption: 97% of respondents currently use or plan to use AI for automating SDLC tasks.
Increased Engineer Demand: As AI accelerates code creation, 76% believe more, not fewer, software engineers will be needed to manage and deploy software.
Human Value: 88% say that essential human qualities like creativity and innovation cannot be replaced by agentic AI.
Career Evolution: 87% believe adopting AI is future-proofing their careers, with 83% expecting significant role changes in the next five years.
Platform Engineering: 85% think agentic AI works best with a platform engineering approach.
Upskilling Required: Nearly half already use more than five AI tools, and 87% wish for greater organizational investment in upskilling.

Compliance and Code Quality Challenges

AI-Driven Compliance Issues: 70% say AI makes compliance management more difficult; 43% see AI compliance/security as a top future skill.
Code Review & Trust: Only 37% trust AI to handle daily work tasks without human review.
AI-Generated Problems: 73% have experienced issues with code produced using natural language prompts (‘vibecoding’) by those lacking strong coding fundamentals.
Technical Debt and Vulnerabilities: Increased use of AI tools is contributing to more vulnerabilities and technical debt, making debugging more complex.
Compliance by Default: 82% predict compliance will be automatically built into code by 2027, yet 76% observe more compliance issues found after deployment than during development.

DevOps Productivity and Collaboration

Frequent Deployments: 82% of organizations deploy software weekly; 60% use more than five tools regularly.
Collaboration Bottlenecks: DevSecOps professionals lose seven hours per week to inefficient processes, mainly due to poor cross-functional communication and disparate tooling.

Industry Perspective

Emilio Salvador, VP of strategy and developer relations at GitLab, emphasizes that organizations must rethink software building and deployment practices in the AI era. While AI amplifies pros and cons, robust DevSecOps processes and team collaboration are essential for harnessing its advantages and mitigating risks.