Engineering leadership and executives are out of touch when it comes to software testing and how AI can be applied to the process, according to Sauce Labs’ 2025 Software Testing Vibe Check: Agentic Edition report.
61% of survey respondents—who consisted of 400 testing executives and engineering leaders surveyed in June 2024—said that their top leadership “lacks a full understanding of software testing requirements.” 65% of engineering leaders agreed with this statement, compared to just 57% of executives.
The report also found that 77% of engineers believe agentic AI will be able to test software autonomously by 2027, compared to only 67% of executives. However, Sauce Labs cautioned that their optimism may not match reality, since most current AI agents fall short of their promises.
“This bullish sentiment reflects the persistent hype swirling about the media and social media platforms. It doesn’t appear to be going away anytime soon,” Sauce Labs wrote in the report.
The company also found 60% of leaders say that when AI makes mistakes, employees take the blame, not the AI providers.
Finally, they found that a majority of respondents are more nervous about giving AI agents full access to company data than they are about falling behind their competitors. 82% of executives feel this way, compared to 63% of engineering leaders.
“The survey reveals a fundamental tension: while organizations recognize agentic AI’s transformative potential, they’re grappling with legitimate concerns about security, accountability, and readiness. The overwhelming preference for hybrid approaches (85% favor mixing human and AI testing) suggests that the future isn’t about replacing human expertise—it’s about augmenting it thoughtfully,” Sauce Labs concluded.
