How AI is Changing Software Development and Testing

It’s no secret that AI has profoundly transformed the way organizations build and test software. How, exactly, is often less clear. Applause’s State of Digital Quality in Functional Testing report for  2026 offers some context: where teams are using AI in development and test, where they’re seeing value from AI, and where human judgment remains critical. Read on for some highlights from this year’s report. 

So far, AI has had a greater impact on development than QA 

In a survey of members of Applause’s global testing community and other software development and QA professionals, 44.2% reported that AI had significantly changed development compared to 35.1% that reported the same level of impact for QA. In addition, 7.9% of respondents reported that AI had no impact on QA, while only 2.5% reported no change to development. 

In-editor coding assistants are the most common use case for AI in development. Integrated AI assistants have become pervasive in most integrated development environments.  

Many organizations still struggle with high defect rates and severity 

While 26.4% of survey respondents reported that both the number and severity of defects have decreased since they integrated AI into the SDLC, 13.2% saw no change and 14.7% saw both the number and severity of defects increase. In addition, 20.3% report that they lack formal documentation around acceptable tools and use cases for AI in development and testing. Balancing speed and quality while adapting to AI-infused workflows was the most common challenge with integrating AI into the SDLC.

One respondent commented. “AI has masked poor performance and reduced quality by degrading raw reporting accuracy… it obscures individual tester skill, and creates a false impression of tester performance over time.”

Human judgment remains critical to quality

Even as teams work to embrace AI, they recognize its limitations. When asked how important human involvement is in functional testing, 86.1% of respondents consider it extremely important and another 13.4% think it’s somewhat important. Fewer than 1% believe human involvement is not at all important in functional testing. 

Human judgment is essential in providing context around expected outcomes, user behavior, and business logic. Providing guardrails for AI is becoming increasingly important as questions arise about the efficacy of benchmarks currently in use to test frontier models. As the technology shifts, regulatory and compliance landscapes are evolving as well – and human judgment is crucial in interpreting complex nuances in these shifting guidelines.  

Read the full report for more insight on AI’s impact on digital quality.