A recruiter in Bangalore told us last year that they had gone from screening 200 CVs manually per week to reviewing 50 AI-shortlisted candidates โ and the quality of interviews had improved because they were seeing fewer obvious mismatches. This is the upside of AI in recruitment. The downside is that the AI doing the screening may be systematically filtering out candidates from certain universities, cities, or demographic backgrounds without anyone realising it. Both stories are true.
What AI Is Doing in Talent Acquisition
CV screening and shortlisting is the most widespread application. Tools like Workday, Greenhouse, and Lever now have AI features that rank candidates against a job description. A recruiter at a company receiving 500 applications for an engineering role sees a ranked list of the top 50, with explanations for why each was ranked.
Automated interview scheduling and communication has essentially eliminated the back-and-forth email chains that used to eat recruiting coordinators' days. AI scheduling tools check both parties' calendars and coordinate without human input. For high-volume hiring at Indian IT services firms and East African tech companies, this is straightforwardly good.
Employee retention prediction is newer and more controversial. ML models trained on HR data can flag employees who show patterns associated with resignation โ changes in communication frequency, performance review trends, tenure compared to peers who left. Used responsibly, this prompts a manager to have a proactive conversation. Used badly, it creates a surveillance dynamic that destroys trust.
The Risks That HR Professionals Need to Own
AI screening tools trained on historical hiring data will replicate historical biases. If a company's successful hires over the past decade were disproportionately from elite institutions, the AI will learn to favour elite institutions. In India, where educational institution bias is already a significant equity issue, and in Africa, where access to top universities is heavily influenced by geography and economic background, this is not an abstract concern โ it will actively hurt good candidates.
HR professionals deploying AI tools have a professional and ethical responsibility to audit these systems for bias, push vendors for transparency about how models are trained, and build human review into any stage where candidates are eliminated rather than just ranked. The efficiency gains from AI hiring tools are not worth building a workforce that is systematically less diverse than your candidate pool.
What HR Professionals Should Build Now
Data literacy โ the ability to interrogate an AI tool's outputs, understand what a "match score" actually means, and spot patterns that suggest bias. You do not need to be a data scientist. You need to be someone who asks the right questions of the data.
AI ethics and audit skills โ knowing how to evaluate an AI vendor's claims about bias testing, what questions to ask during procurement, and how to design human review processes that catch what the algorithm misses.
People skills, redefined โ as AI handles more of the transactional parts of HR, the human work becomes more valuable: complex conversations about performance, culture, career development, and wellbeing. The HR professionals who are excellent at these conversations are more differentiated than ever.
Data skills for HR professionals
Understanding how AI models work โ and where they fail โ is increasingly essential for HR leaders evaluating and deploying these tools. The Data Science and Data Analysis tracks give you that foundation.
Explore the Data Science Track โ