Cambridge, UK — June 09, 2026 — New research from Gearset, the leading enterprise Salesforce DevOps platform, reveals that enterprise software teams aren't approaching AI adoption uniformly. Trust in AI and autonomous agents varies depending on where teams are in the software delivery lifecycle (SDLC). The findings come from the State of Salesforce DevOps 2026 report.
Trust is highest where AI delivers the most value with the least risk. At the build stage, where developing changes is time-intensive and mistakes can be caught before causing issues, 82% of teams trust AI involvement. Trust remains strong through validation and observability, stages where AI can accelerate quality checks and surface issues without touching live systems. Notably, teams also show strong appetite for autonomous agents specifically around planning and observability, where the potential for efficiency gains is high and the consequences of errors are manageable.
Trust drops sharply at release and operate. Only 58% of teams trust AI at the release stage, with 42% excluding it entirely. This isn't a rejection of AI, it reflects sound engineering judgement. Release and operate are where changes land in live production environments that the business depends on, and where deterministic CI/CD pipelines and automated backup already solve the core problems extremely well. Teams see little reason to introduce AI where proven automation handles the job, and where the cost of failure is greatest.
What makes this picture particularly striking is how teams are treating the AI-generated changes that do make it through the pipeline. Gearset's research finds that 43% of enterprise teams review AI-generated code and configuration in exactly the same way as human-written changes, with a further 33% applying even greater scrutiny. That level of equivalence, holding AI output to the same bar as a developer's work, reflects just how embedded AI has become in the day-to-day delivery process.
Gearset's CEO, Kevin Boyle, said: "What our data shows is that enterprise teams have developed a clear-eyed view of where AI earns its place in the delivery pipeline. They're embracing it enthusiastically for building, planning, validating and observing, exactly where it can save the most time without adding risk. And they're being appropriately cautious at release, where deterministic automation already works and the stakes are highest."
To read Gearset's State of Salesforce DevOps Report 2026, click here.
