Jane Edwards 
- Tricentis has found that 60 percent of organizations deploy untested code amid increased AI adoption
- Poor software quality results in annual losses exceeding $1 million for 20 percent of organizations
- At least 80 percent of organizations trust agentic AI to make release decisions
Tricentis has released its 2026 Quality Transformation Report, which found that 60 percent of organizations deploy untested code to production as artificial intelligence adoption accelerates software development.
In a June 3 news release, Tricentis said the report is based on a survey of more than 2,500 CEOs, chief information officers, chief technology officers, engineering leaders, software developers, and DevOps and quality assurance professionals across industries, including the public sector, energy and utilities, manufacturing, retail and financial services.
Why Are Organizations Still Deploying Untested Code?
According to the report, six in 10 organizations said they move untested code to production, a figure that remains largely unchanged from 2025.
The report found that organizations previously attributed untested code deployments to accidental quality issues. In this year’s report, respondents said they deploy untested code primarily due to leadership pressure to prioritize speed over quality and the growing volume of AI-generated code that teams struggle to test fully.
Tricentis CEO Kevin Thompson said AI could help software development teams move faster, but organizations face greater risks when software quality processes do not keep pace with the speed of development. He also noted that organizations often take shortcuts that reduce confidence in software quality.
“Our research highlights the growing pressure teams are facing to balance speed, quality and control as software development accelerates. As risks like financial performance and customer trust become more visible and measurable, software quality can no longer be treated as just an engineering concern. It must become a boardroom imperative,” Thompson added.
Are Organizations Ready to Govern Agentic AI at Scale?
The report found that 83 percent of organizations trust agentic AI to help guide release decisions, while 82 percent reported readiness to implement and govern AI agents at scale.
However, the respondents also reported ongoing challenges, including untested code deployments, security concerns, tool sprawl, skills gaps and data quality issues.
How Much Is Poor Software Quality Costing Organizations?
According to the report, one in five organizations said they lose more than $1 million annually because of poor software quality.
The respondents identified security and compliance failures as the leading cause of those losses, followed by technical debt and rework. Nearly half of organizations estimated annual losses ranging from $500,000 to $1 million.
How Do the Findings Align With Tricentis’ Push for Automated Testing?
The findings align with Tricentis’ ongoing emphasis on automated testing as AI accelerates software development.
In a commentary, Benjamin Baldi, senior vice president of global public sector at Tricentis, called on defense agencies to transition from manual testing to AI-powered automation to accelerate the adoption of new capabilities while maintaining software quality, security and compliance with the Department of War’s Acquisition Transformation Strategy.
The company’s 2025 Quality Transformation Report similarly highlighted the need for scalable automated testing approaches to strengthen software development and delivery across the public sector.
