Artificial Intelligence in recruitment needs to get better

Artificial Intelligence is rapidly transforming the business landscape, improving efficiency and speed. AI applications are available across various business segments, and the option to customise them for individual firms’ needs further enhances their acceptance and popularity.

HR is an area where AI has proven highly useful. The power of AI is being leveraged in various HR functions, including recruitment. Recruitment is a core HR activity, followed by onboarding, training, and development.

Recruitment is usually a time-consuming process. In a country like India, where a single vacancy can attract thousands of applications, the challenge of shortlisting candidates and scheduling the selection process is significant. Thankfully, the trend of physical applications is coming to an end, and AI is now handling e-applications and resumes. 

In our country, private-sector companies and start-ups have taken the lead in AI-based recruitment. Also, several companies have emerged to provide platforms and technology for this.

Companies find many advantages in using AI to meet their manpower needs. Whomever we talked to mentioned time saving as the biggest advantage. Using AI in recruitment means delegating administrative tasks to intelligent tools—automating routine tasks, such as resume screening and scheduling, results in substantial time savings. With AI’s ability to analyse and process large datasets, firms can identify pockets where suitable candidates are found and predict candidate success based on historical data. These pockets include social media platforms and professional network sites. Specific AI tools can schedule interviews, send follow-up emails, and track all recruitment activities.

The work begins with predictive analytics to analyse past hiring data and use AI to forecast future staffing needs, enabling proactive recruitment strategies and workforce planning. Business expansion is also taken into account for assessing manpower requirements.

Carefully designed AI systems can provide an objective evaluation of a candidate by targeting specific skills, experience and other predefined criteria. Use of these systems is expected to give way to a more meritocratic and objective process.  

We can continue to list the advantages of using AI in recruitment, but can we rely entirely on it and consider it foolproof? It is still a work in progress, and it has its own pitfalls. Continuous evaluation of the process and results, adding intelligence to the tools to make them more sophisticated, and investments in other areas promise a bright future for AI-based recruitment.

AI in recruitment processes handles a large volume of candidates’ personal data, which may contain sensitive information. If a data breach occurs, it may result in significant financial and reputational damage to the firm. Therefore, strict compliance with data protection regulations and robust cybersecurity measures should be implemented.

We should remember that, while we may be using machines and machine learning in recruitment, we’re not recruiting machines or robots. Humanising AI is a significant challenge, and it also applies to AI in recruitment.

AI algorithms are meant to learn from historical data, which might contain unconscious human biases. If not handled properly, AI tools may perpetuate these biases. This will lead to discriminatory hiring practices and unethical hiring decisions.

The AI engines are trained on years of historical data that relies on firms hiring more men than women. The algorithm incorporates the bias and scales it across the entire candidate pool. Other issues include preferring candidates from certain institutes (e.g., top IIMs), penalising study and career gaps, which are more common among women, and favouring candidates whose resumes are drafted in a particular way. Human recruiters also make mistakes, but there is a difference. When a human recruiter makes a biased decision, it affects one or two candidates; when an AI makes that decision, a large number of candidates are affected systematically and invisibly.

AI in recruitment relies on pattern matching, which should also be subject to scrutiny. A candidate who scored well in 10+2, joined a top engineering college, got campus selection in a leading analytics firm, will be preferable as a lateral hire than the one who after becoming an engineering graduate remained jobless for a few months, then joined a consultancy; while working taught herself advanced analytics and now wants to make a career switch in a related role. This candidate’s motivation may be higher, but we can’t be sure whether AI will recognise it.

The very diversity that firms want to bring into their HR may get ignored in AI recruitment. AI is efficient at information processing, whereas humans are good at interpreting context. AI can find correlations in available data, but it isn’t equipped to understand the causal relationship and contextual nuances that make someone a better hire.

AI is a modern tool, but it is more inclined towards conformity than discretion. In recruitment, it may judge candidates based on past performance but fail to recognise their potential. AI-based recruitment should also be studied from the angle of candidates’ experience. For some, an automated tool may be less threatening, but automated rejections without feedback or encouragement, mechanical interviews without a human interviewer, and questions that read more like interrogation than conversation may not be a pleasant experience, even for promising candidates. Firms may be using AI to improve efficiency in their HR processes while overlooking factors that attract top candidates, such as a sense of value and understanding.

Despite its limitations, AI in recruitment will be used increasingly. We can see it improving and hope it becomes more reliable. At least for roles where the stakes are higher, it should be used with extra care and alongside in-person assessment and evaluation. A hybrid model is always better than a standalone model. The cost savings from AI-based recruitment shouldn’t come at the expense of a wrong hire.

(The author is based in Mumbai)