Artificial Intelligence: The Present and Future of Human Resources Recruitment and Selection Processes ^†

Abstract

Artificial Intelligence (AI) is a broad term that usually refers to a diverse set of computational procedures that can mimic human decisions and/or processes so closely that they appear intelligent, being able, for example, to process quickly large volumes of data. AI is such a powerful tool that organisations are increasingly using it in various areas, including human resources (HR) management, especially in recruitment and selection functions. For instance, big data algorithms are highly instrumental in expanding the process of searching for candidates. However, there are several key questions that are still unresolved regarding ethical issues and the reactions and attitudes towards AI of its users (recruiters, selection managers and potential candidates), necessitating a more extensive empirical and systematic review of the literature at this level. In this context, this paper discusses AI and its applications in HR recruitment and selection processes, addressing the future trends and challenges defined in the existing literature.

1. Introduction

The world of work in the 21st century is strongly linked to the incorporation of information technologies and its multiple variants and applications, and it has been pointed out that we are witnessing a fourth industrial revolution [1]. The term fourth industrial revolution was coined by Klaus Schwab [2], founder and executive chairman of the World Economic Forum, to describe the exponential development of the digital revolution (that has been taking shape since the middle of the last century), characterised by a fusion of technologies that blurs the lines between the physical, the digital and the biological [3]. Among all the relevant technological forces in this fourth industrial revolution (e.g., 3D printing, quantum computing, nanotechnology, biotechnology, alternative forms of energy technology, and so on) [4], artificial intelligence (hereafter AI) stands out as an emerging and powerful technology that has received much attention in the popular press, academic research, and industry.

AI is a broad term, coined by John McCarthy in 1955, that usually refers to a diverse set of computational procedures that can mimic human decisions and/or processes so closely that they appear intelligent, being able, for example, to process quickly large volumes of data to identify, relate and predict patterns [5]. AI is such a powerful tool that organisations are increasingly using it or considering its use in various areas. Moreover, 85% of executives surveyed in a global study are projected to invest heavily in AI technologies within the next three years [6]. Furthermore, it has been predicted that this technology would significantly change the business landscape in the 21st century [7]. Numerous white papers and research reports have presented the benefits and advantages of implementing AI, claiming that AI could change organisations, industries, and society in the future [8]. In sum, AI can be regarded as a disruptive technology, as it will fundamentally reshape our lives and work [9].

AI has evolved significantly in recent years. There is, however, a lack of complete and comprehensive understanding of its use, impact, influence, and critical success factors in organisations [10], specifically regarding human resources management (hereafter HRM) and more concretely in HR recruitment and selection. More than half of the companies using AI do so precisely to improve and optimise such recruitment and hiring processes [11,12]. AI helps organisations with the reduction of costs, especially in terms of time, effort and in repeating daily tasks for recruiters, as well as for candidates in reducing the organisations’ response time, favouring a positive employer brand [13]. However, the use of AI for HR recruitment and selection purposes is not without controversy and criticism. AI might also have a negative impact, affecting its use to diversity management in organisations [14] and even leading to the substitution or replacement of humans in certain types of tasks and jobs [15]. Moreover, there are several key unresolved questions regarding ethical issues and the reactions and attitudes towards AI of its users (recruiters, selection managers and potential candidates) [1], necessitating a more extensive empirical and systematic review of the literature at this level.

In this context, this paper discusses AI and its applications in HR recruitment and selection processes, addressing the future trends and challenges defined in the existing literature.

2. AI Applications in HR Recruitment and Selection Processes

AI was defined in general terms as a diverse set of computational procedures that can mimic human decisions and/or processes so closely that they appear intelligent, being able, for example, to process quickly large volumes of data and to identify, relate and predict patterns [5]. In other words, “AI has the ability to make decisions in real time based on pre-installed algorithms and computing technologies constructed based on data analysis to learn and acclimate automatically to offer more refined responses to situations” [16] (p. 1).This description shows its potential, for example, in terms of data processing (e.g., sourcing and refining) and decision making [17,18], two tasks or processes that are usually present in HR recruitment and selection processes, meaning that AI is increasingly being used by organisations.

Wisskirchen et al. [5] described five main applications of AI: (1) machine learning (hereafter ML) (or the use of machine and computer programming to optimise a certain performance criterion using example data or past experience [19]); (2) robotics (the utilization of machines capable of performing automatic tasks or simulating human behaviour); (3) dematerialisation (or the transformation of traditionally physical products into software); (4) the gig economy (or working on platforms, whether working in teams or on demand through applications); and (5) autonomous driving (or vehicles that are capable of self-driving using sensors). Among all these possible applications or uses of AI in HR recruitment and selection processes, ML and robotics are highlighted here.

ML can be distinguished from deep learning (hereafter DL), a more advanced form of ML. Both are based on a set of algorithms that try to model high-level abstractions in data, but the main difference between the two is that, in DL, the algorithms are based on artificial neural networks aimed at making the machine learn on its own [19]. The algorithms contained in ML make it possible for computers to conduct specific tasks autonomously (i.e., without the need to be programmed) that mainly have to do with identifying patterns in large data sets to make predictions. For example, techniques such as natural language processing (NLP) can collect and analyse data sources in an automated and rapid way [20]. These techniques are therefore an essential part of big data, and although it is not unusual to understand them as synonymous with so-called data mining, it is their ability to reproduce patterns and make predictions based on them that distinguishes them from data mining (which is more exploratory and descriptive in nature). Depending on the goal pursued when using ML/DL, it is possible to distinguish several types of algorithms that can be used such as supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning [21]. As Rogers et al. [16] noted, detailed key cases of ML and DL in an HR recruitment and selection context might include anomaly detection, background verification, content personalization, as well as questions of ethics and data management for HRM practitioners, such as images, video, and speech recognition.

Robotics can be described in general terms as the use of robotic machines that have the capacity (depending on their software that can include ML/DL [22]) to perform automatic tasks or even simulate human behaviour. It is important to distinguish between bots (or computer programmes that include rules executed repetitively on the internet, which allow them to perform certain actions autonomously), co-bots (or collaborative bots) and chatbots (or specialised bots created to carry on conversations and provide preconceived responses). AI applications for HR recruitment and selection purposes usually involve the use of tools such as bots and chatbots. For example, bots can define quickly the most suitable profiles for a specific position, shortening the time frame, for example, in the pre-selection phase. Chatbots can initiate a real time communication with applicants for the job in the form of a screening interview, reducing the organisations’ response time and improving the experience of candidates [23]. Fraij and László [24] summarised in their recent review some of the chatbots used by big companies such as Ikea or Amazon, for example, Xor (https://xor.ai/ (accessed on 1 August 2023)) and Talkpush (http://www.talkpush.com/ (accessed on 1 August 2023)). In short, AI offers techniques and tools that can support and optimise several tasks in the HR recruitment and selection process [25].

Table 1. HR recruitment and selection process: tasks and AI applications (adapted from Laurim et al. [25]).

HR Process	Tasks	IA Application
Recruitment	Job advertisement	ML and NLP can help recruiters by identifying the keywords that the advertisement should contain and the right channels for its communication.
	Job application	Intelligent digital assistants (often based on DL) can take over the task of writing the job application for the applicant. This allows all data to be available online, so that such intelligent digital assistants can collect, prepare, and send the data to the applicant management system. The competences and skills identified can furthermore be compiled and verified (e.g., via Open Badges).
	Organisation response Communication	Self-learning chatbots capable of interacting in a way that can answer frequently asked questions from applicants and guide them in some directions.
Selection	Pre-selection/evaluation	Automated pre-selection (depending on the programming of its software) can recognise both hard and soft skills and even personality traits that are not clear in the candidate’s CV or presentation videos. Preliminary interviews through chatbots

summarises some of the main applications of AI in such HR recruitment and selection tasks.

3. AI Trends and Challenges in Recruitment and Selection Processes

The use of AI and its applications in HR recruitment and selection has numerous advantages. These include more reduced costs in terms of time, effort, money, and human resources [13,26]. HR professionals can focus on those tasks that have more to do with reasoning, reviewing and monitoring the process, making decisions, and interacting and communicating with candidates. AI brings greater accessibility to numerous forms of data (e.g., large volumes of résumés and social media content) that were previously burdensome and more difficult to analyse, due to its power to manage enormous quantities of data [26]. In doing so, another advantage of AI is its potential to increase talent [27] and diversity [28] within the organisation, as well as to provide predictive accuracy [26]. On the one hand, recruitment may be enhanced using ML, as noted above, in the first stages of the recruitment process. On the other hand, its use can drop potential biases in the selection process (related, for example, to characteristics such as gender and/or age).

However, these potential advantages may also be some of AI’s main risks and drawbacks, becoming challenges to cope and trends for the future. Gónzalez et al. [26] summarised the actual potential limitations of AI and its applications, such as in ML, in four categories: data quality, “black box” predictions, ethical and legal issues and users’ reactions (recruiters, selection managers and potential candidates) towards AI. For example, it has been pointed out that the use of AI as a decision support element may lead HR managers to perceive that they have less autonomy in their jobs and that the selection decision was already predetermined, thus becoming less accountable for their own decision making [29]. Moreover, AI learning databases can be biased, reproducing and maintaining initially discriminatory algorithms, and the unconscious biases and stereotypes of AI designers may also cause discriminatory behaviours [14] in its application for HR recruitment and selection purposes. In this vein, some recent proposals such as Rodgers et al.’s [16] could be useful. The authors proposed a model of HR selection decision making using AI, which shows how employing certain strategies to build the algorithms (e.g., creating a diversified multi-stakeholder expert advisory board, respecting differences of opinion, to ethically guide the AI) can support ethical AI use.

It will become increasingly necessary in HR recruitment and selection processes to collaborate with automated systems. Therefore, it is necessary to adopt broader frameworks of analysis that consider how the changes arising from the use of AI can be realised with a view to designing and implementing such systems in the optimal way for organisations as well as for individuals. Progress in this direction is needed not only at an empirical level, but also in strengthening the collaboration among HR professionals, AI designers, electronic scientists, legal scholars, and members of other professional disciplines important in the development, implementation, and evaluation of AI applications in organisational contexts.

4. Conclusions

This paper discussed AI and its applications in HR recruitment and selection processes, addressing the future trends and challenges defined in the existing literature. AI and its applications for HR recruitment and selection purposes (e.g., ML, DL, PNL, bots, and chatbots) have evolved significantly in recent years. Moreover, its use will grow due to its potential advantages (e.g., more reduced costs in terms of time, effort, money, and human resources [13,26]). However, these potential advantages may also be some of AI’s main risks and drawbacks, becoming challenges to cope with and trends for the future (pertaining to, e.g., ethical and legal issues, diversity, equity and inclusion (DE & I), and users’ reactions towards AI). Therefore, it is needed to progress not only at an empirical level, but also in strengthening the collaboration among HR professionals, AI designers, electronic scientists, legal scholars, and members of other professional disciplines important in the development, implementation, and evaluation of AI applications in organisational contexts.