What is the lever for a successful integration?

The integration of artificial intelligence into HR activities and decisions is a lever of value and a competitive advantage.

A Deloitte survey, for example, reveals that 22% of top performers manage part of their hiring process automatically – compared to 6% of lower performers. In the sense that it makes it possible for candidates to be selected faster, more efficiently and less bothered by human bias, AI thus makes it possible to generate new revenue and reduce the cost of HR functions.

However, it remains a matter of debate: on the one hand, for employers and other HR decision makers, who are surprised about their autonomy, and on the other, for candidates. Like any change, adopting AI brings with it legitimate concerns. Also, instead of blaming or clinging to disobedience, it is better to understand the reasons for this fear and give a clear answer, so that everyone can AI in the recruitment process. An informed use of it may be allowed. To this end, the response of candidates to the use of AI in the selection of scientific knowledge – positive or negative – reflects the impact of various relevant factors, in particular: (1) its position in the recruitment process, (2) awareness, (3) human supervision.

The beginning or the end of the process?

The potential for candidates to enhance their professional skills in the selection process is an important factor in realizing its fairness; Naturally, a job of recruitment is considered fair when people get a chance to show all their skills before making a decision. This observation raises questions about the place of AI in the whole process: is it more relevant to place it at the beginning or end of the process to give candidates a better chance of demonstrating their skills and thus increase perception fairness? Method?

Although this positioning may vary depending on the type of position legally – and fortunately – filled (rarity and deficit of profile for recruitment, number of applications, etc.), it appears that when AI is used as selection at the beginning of the process, at the end of the process Without being a decision-making tool, the perception of fairness is enhanced. Recently, Anna Lena Hankenscrower and Christoph Lutge, researchers in business ethics at the University of Munich, observed that an interview using AI techniques was considered more beautiful when used in the preliminary election round. Similarly, if we are interested in other forms of AI, such as predictive recruitment tools based on the use of psychometric testing, their positioning in start-up processes may seem as relevant as possible to benefit from their full potential. , But also to meet the needs of candidates for justice. The use of psychometric tests allows candidates to show aspects of themselves that cannot be inferred from the CV, such as their personality, their motivation and values ​​or their thinking. Beyond allowing for fair and more efficient recruitment (behavioral data further predicts an individual’s future success), integration of such tools at the beginning of the process will make it easier for candidates to gain acceptance for its use.

Need to explain

No matter where the AI ​​is in the recruitment process, candidates need to understand why the tool is being used and for what purpose. According to some studies, candidates’ feedback can thus be positively influenced by raising awareness about AI: the information provided can reduce uncertainty, increase transparency and improve the credibility of the selection process. Candidates, for example, find AI more fair if they are made aware of the possibility of reducing human bias in their selection process: the perception of fairness is then linked to aspects of appropriation, such as freedom prejudice, or equal opportunity for all applicants. Similarly, increasing the effort to explain will increase the acceptability of other tools, such as psychometric testing in the selection process. Recently, Klaus Melchers and Barbara Corner of the Universities of Ulm and Zurich have shown that providing test explanations significantly increases the perception of the fairness of the process among candidates. Other features, such as access to detailed feedback after the exam, or the possibility of carrying them online, also help increase the acceptance of candidates.

Human supervision

The extent to which AI is integrated in the final hiring decision varies from company to company. In some cases, AI does more work, providing additional information to HR decision makers and analyzing candidates … but employers are in control of the final decision. Among others, AI has already adopted an automated decision-making process, which includes verifying or rejecting applications. Even though over time, strategies have been developed for the acceptance and use of AI, there is now some empirical evidence that demonstrates the value of human supervision. On the one hand, decision makers thus prefer to rely on AI if they have the potential to adjust their decisions. In her case study, Elmira van den Brock, a researcher in artificial intelligence at the University of Amsterdam, identifies an overall choice of human resource managers who are able to make exceptions and adjust the decisions made by AI according to context. For them, the ability to distinguish between temporal changes in context and supply and demand is important in design process selection. On the other hand, for candidates who expect companies to invest time and labor in the selection process, it is not wise to set up a fully automated recruitment process at present: for example, candidates find automated interviews to be fair. When the final decision on hiring under human supervision is made.

In the case of recruitment, AI has demonstrated its effectiveness mainly: for example, a meta-analysis comparing clinical and mechanical decision-making in the selection process shows that in order to predict job success, a simple equation proves to be more efficient and precise. So the question now is whether its interest in the HR decision is legitimate, but how it can be integrated to democratize its surplus value and maximize its acceptability, especially to candidates.

Leave a Comment