Untested code is still rampant: 60% of organizations continue to ship untested code into production, with teams now knowingly doing so due to top-down leadership pressure (32%) or having too much code to test (30%).
A costly trust gap exists between leadership and practitioners: 93% of C-level respondents feel confident in their testing strategies, while nearly one-third of QA and DevOps leaders are uncertain or unconfident — and only 56% of QA/DevOps professionals trust the AI-driven tools informing delivery decisions, versus 81% of CEOs.
Governance of AI-driven delivery is critically underprepared: While 82% of organizations feel at least somewhat ready to govern AI agents and autonomous workflows, only 35% feel fully prepared to manage those environments at scale.
Operationalizing quality and trust in the era of agentic software delivery
The past year has marked one of the most supercharged periods in the history of modern software development. Unlike anything before, AI acceleration has transformed every stage of the
software development lifecycle (SDLC), fundamentally changing how organizations build, test, release and manage software.
As uncovered in Tricentis’ second annual Quality Transformation Report, there is a growing tension at the heart of modern software delivery. Organizations are under mounting pressure to move faster in an AI world, yet many lack the operational trust, alignment, and oversight needed to scale software quality confidently, and at AI speed.
The Tricentis 2026 Quality Transformation Report endeavors to uncover what has changed since the inaugural report in 2025, as well as explore how modern organizations are using AI differently than they did a year ago.
This year’s findings showcase persistent concerns about shipping untested code, the strains that AI is putting on software quality systems, an increasing challenge related to operational trust within organizations, and the growing importance of governance on agentic software delivery.