IBM’s AskHR handled more than 16 million employee messages in 2025. The scale is striking, but the company’s account also records a difficult start. After IBM directed employees to the service, its internal Net Promoter Score (NPS) fell to minus 35. IBM refined the system and improved access to human experts; it now reports a customer satisfaction score of plus 74, a different measure from NPS.
Indian employers face a higher-stakes version of the same problem when AI moves from answering questions to screening job applicants. Speed is visible. Selection quality takes longer to establish and requires a record of who advanced, who did not, and why.
HR leaders call for a useful division of labor. Software can handle administrative volume. Recruiters remain responsible for interpreting evidence, reaching overlooked candidates, and correcting decisions that do not stand up to review.
Faster Screening Requires a Separate Test of Quality
IBM’s results concern employee support, not recruitment. They show that software can process requests at scale, but they do not show whether an algorithm identifies stronger candidates. That distinction should shape the business case for any hiring tool.
Scheduling an interview follows an agreed procedure. Ranking applicants requires decisions about which qualifications matter and what counts as evidence of ability. Processing more applications is valuable only when those decisions can be explained and checked.
Murali Santhanam, CHRO at AscentHR Technologies, sees an opportunity to redirect recruiters toward harder sourcing work. “Recruitment teams should start focusing on non-traditional sourcing and hiring… creating a database of known candidates, improving employee referral candidate acquisition, and head hunting with a targeted effort.”
Sonica Aron, founder and managing partner at Marching Sheep, describes the division this way: “AI will take care of the volume, while the HR will be responsible for taking care of empathy and context.”
Hardeep Singh, president at Right Management India, focuses on the decisions that follow processing. “The outcome is not just efficiency, it’s clarity. HR teams are spending more time interpreting data, engaging with the right talent, and making better-informed decisions.”
Those gains depend on how the released time is used. A shorter queue can give recruiters more room to examine uncertain cases and reach candidates who conventional sourcing misses. If review capacity falls at the same time, the organization needs separate evidence that selection has improved.
“The outcome is not just efficiency, it’s clarity. HR teams are spending more time interpreting data, engaging with the right talent, and making better-informed decisions.”
— Hardeep Singh, president at Right Management India
India’s referral practices show why processing and selection need different measures. A referral can bring a candidate to a recruiter’s attention without proving suitability for the role.
Santhanam points to the time saved in reviewing applications. “The time it takes to browse through the number of profiles, screens and shortlists is significantly reduced to minutes using AI agents and AI applications to provide the best-fit versus requirement.”
Aron expects referrals to face a more explicit assessment. “Referrals will no longer be considered just as ‘word of mouth’ endorsements; their skills and potential will be assessed by the AI system to see if they match the profile of a job.”
“Referrals may open the door, but they no longer determine who walks through it,” Singh says.
Employers can test that proposition. They can compare referred and independently sourced applicants against the same published requirements, then review who advances and why. A consistent procedure helps, but it does not prove that the requirements or their weighting are appropriate.
Language Capability Does Not Settle Cultural Interpretation
Language coverage matters in a country where employers recruit across regions, educational backgrounds, and levels of English fluency. It does not answer the harder question of how a system interprets an applicant’s response.