Using AI in software development and deployment has fast become an essential part of enterprise teams’ workflows. Google Cloud’s 2025 DORA report found that 90% of software development professionals use AI in some capacity, transforming the day-to-day for engineers who would have been occupied primarily with writing and testing code before. But under the surface, teams aren’t ripping up the rule book to build trust in AI-generated code – they’re treating AI like any other teammate.
As MIT’s Computer Science and Artificial Intelligence Laboratory reported, AI coding’s widespread adoption doesn’t mean it can handle all aspects of large-scale software engineering by itself. Releasing to production is higher stakes than writing code, for example. Scrutiny from real humans is vital to build trust further down the software lifecycle. And despite some early teething issues, like Amazon’s coding assistant-related outage, early signs are that enterprises are learning how to safely welcome AI to the team. Teams with effective processes aren’t replacing their employees with AI, they’re amplifying human ingenuity to produce higher quality code at greater speeds than ever.
Treating AI as a capable new hire
Our own research has found that teams are learning how to trust AI. Almost half (46%) of teams are confident in AI-generated code performance, indicating that success comes down to how AI is governed, not how effective the tooling is.
Additionally, 43% of enterprise teams review AI-generated code and configuration changes in exactly the same way as human-written work, while 33% apply even greater scrutiny. This tallies with other industry studies looking at how teams are learning to embed AI into their existing workflows. McKinsey gave the example of AI taking the heavy coding work overnight at many teams, while engineers apply judgment to its outputs during the ‘day shift’. In both cases presented here, AI output is being treated in the same manner as a human teammate.
This tells us a lot about how mature teams are approaching AI. To make AI adoption an iterative process that improves over time, starting with low-stakes, repeatable tasks and applying stringent checks to outputs gives the tech a chance to work properly without expecting immediate results.
AI adoption does require a big shift in how teams interact with the tools they use, but the fundamentals that ensure success aren’t changing. Most businesses are practicing responsible AI adoption rather than allowing changes to bypass established controls.
Trust is a learning process
Trust in AI is not equal across the software development lifecycle. The vast majority (82%) of teams we surveyed now use AI during the build stage, dropping to 58% at release where production risk is highest.
Using AI primarily for earlier stages of software development is a rational step to make sure failures don’t impact the wider business. Reviews, automated testing and deployment controls provide confidence that mistakes can be identified before they affect customers.
Outputs are much easier to review and refine before the release stage and with many businesses still lacking full observability, teams often only hear about mistakes in live code once users notify them. Despite claims that software development could eventually be fully automated, teams are showing a clear awareness of where human oversight remains essential. Rather than making blanket judgments about AI’s capabilities, they are taking a more nuanced view, which bodes well for the future of AI-assisted software delivery.
Scale with guardrails
There is continuity in how mature teams are making sure AI-generated code is fit for purpose, but it’s still having a seismic impact on how software products get delivered. Developers have far more free time as a result of their new virtual teammates, so the value they provide is moving up the stack. Less time is spent producing code from scratch, and more on decision-making about how software should evolve.
AI shifts the cognitive load from execution to judgement. While it can accelerate implementation, teams still need people who understand the business context behind a product being built by AI. As the Google DORA report argued, AI amplifies existing organisational strengths and weaknesses rather than replacing them. Human decision-making and reviews are what make AI adoption pay off at enterprise scale.
As AI takes on more of the mechanical aspects of software development, accountability becomes increasingly important. If someone asks why a feature was built a certain way, teams still need people who can explain the reasoning behind the decision. AI can generate an answer, but it cannot own the outcome.
The organisations that benefit most from AI will not be those with more automation, they will be the ones that combine AI-driven speed with engineering discipline and a strict adherence to repeatable processes.
AI adoption is a DevOps challenge
The debate around AI-generated code often focuses on whether the technology can be trusted. But industry reports show that the best teams are building trust over time by doubling down on processes. As AI capabilities continue to improve, the gap between high-performing teams and everyone else is unlikely to be determined by access to the latest model. The engineers and teams that will thrive are the ones who pair AI’s capabilities with their own judgement, and the data suggests many are doing exactly that.
