Evolve or perish: Agentic AI set to disrupt software testing roles

The Indian IT sector currently has over 375,000 active professionals in testing and QA/QC functions across experience levels, according to data sourced from specialist staffing firm Xpheno

Artificial intelligence (AI) and agentic AI technologies are triggering a major shakeup in software testing, one of the most traditionally structured functions within the software development life cycle (SDLC).

 

With AI-driven automation increasingly replacing repetitive and rule-based testing tasks, quality assurance (QA) and quality control (QC) roles stand the risk of becoming obsolete if they do not evolve.

 

The Indian information technology (IT) sector currently has over 375,000 active professionals in testing and QA/QC functions across experience levels, according to data sourced from specialist staffing firm Xpheno.

 

This talent pool, however, has remained largely stagnant, with net growth of under one per cent in the past year (until April 30). The negligible net growth can be attributed to the companies adding only entry-level and low-mid level employees — those with up to six years of experience. The talent pool at the middle level and above has marginally reduced or remained stagnant. 

 

“The contraction is primarily an outcome of natural attrition not being entirely refilled by enterprises. Overall, the pause is hiring for testing and QA/QC functions in the tech sector seems to be a tactical move to allow ‘wait-and-watch’ on AI intervention and impact,” Kamal Karanth, cofounder of Xpheno, told Business Standard.

 

But not all testing jobs are at risk, according to experts. It is mainly the lower end of the function, comprising about 40 per cent of the total talent pool, which is repeatable and transferable. The need for entry-level talent in this capability layer will gradually reduce as agentic AI tools and processes mature in the years to come. Roles like test engineers, application testers, QA testers, software test engineers, and QA engineers, would then become replaceable.

 

“It is likely that this very profitable business will be thoroughly disrupted over the next 2-5 years. The testing function will be dominated by AI model companies and integrated into the broader AI coding platforms,” said Peter Bendor-Samuel, founder and executive chairman, Everest Group. 

 

Most IT services players do not report QA and QC as a separate revenue line, but this segment is a significant contributor to the top line. Most of the application, maintenance and development (AMD) — still a good chunk of revenue for the sector — includes embedded QA or QC services.

 

“Testing has always been a low-hanging fruit on the automation tree, and one of the most automated aspects of SDLC,” said Hansa Iyengar, practice leader of BFS and IT services, HfS Research. 

 

That implies popular automation testing tools like Selenium, Cypress, Playwright, and Appium are freeing up a lot of time for these engineers, getting them to think about “test strategy” rather than “fixing broken scripts”, according to a recent blog on IT industry body Nasscom’s website.

 

A lot of testing tools are slashing testing time, boosting coverage, and enabling real-time quality checks across development cycles. Agentic AI is taking it a step further with self-healing scripts and autonomous test agents. These can now adapt to code changes without human intervention, making testing more resilient and scalable. 

 

“For testers, this shift is less about replacement and more about reinvention. As routine tasks get automated, the testing role is evolving into one that demands oversight, judgement, and collaboration with AI systems. Testers will focus on complex scenarios, AI output validation, and shaping test strategies that align with business goals,” Iyengar added.

 

From a talent standpoint, reskilling is the only option to go up the value chain. While legacy roles will wither, newer roles in testing such as architects, specialists, analysts, consultants, and governance professionals, will become far more prevalent.

 

“AI is generating even test data by itself, but the roles that are not getting impacted are those that create a platform or tools,” said TeamLease Digital Chief Executive Officer Neeti Sharma. “There will be disruption in the short term, but the skillsets will change, and high-value testing work will grow.”