Can you accurately hire by removing human interaction? Launched in November 2022, ChatGPT took the world by storm. The advanced natural language processing model, developed by OpenAI, distinguished itself from previous artificial intelligence (AI) tools by providing realistic human-like conversations. Simply, you ask it a question and it answers you as if it were a human.
Based on its early uses, ChatGPT could have a serious upside for recruitment – to the extent that it has the potential to play a pivotal role in a firm’s search for talent and its use has become increasingly prevalent in the world of executive search[1]. Firms embracing change cite they have demonstrated a more diverse and potentially higher quality pool of talent. Perhaps unsurprisingly, HR is not the only sector pushing to use AI platforms. In fact, advances in AI are expected to lead to widespread economic and social changes, with nearly 14% of the global workforce needing to change occupation by 2023[2].
At Dryden, we are engaged by firms to find the best executive search talent. We do not currently employ AI. So the question we need to ask is whether the good of AI outweighs the bad.
What are the benefits?
There are three key benefits of introducing AI into a search process: it saves time for recruiters; it matches candidates better, based on their skills, to the jobs available; and it removes bias.
1. Automating processes
AI in recruiting enables search firms to connect with the best candidates without spending vast amounts of time reading CVs[1]. AI algorithms can analyse CVs, saving recruiters’ time and energy. This frees up time for more important things like developing long-term relationships with candidates, selecting candidates for positions, and setting up interviews[2]. ChatGPT can also be used to build talent pipelines by automatically re-engaging with passive candidates, which further cuts down a recruiters heavy lifting,[3] and is proving a key reason to adopt the technology. Five out of eight professionals say they would use AI in the recruitment process if it saved time[4].
2. Candidate experience
Recruiting has complexities outside of just finding the most suitable CV. ChatGPT can help generate emails, provide information about roles, interview tips and feedback on practice interviews. This constant source of information and feedback creates a more engaging experience for candidates[5]. This is beneficial as 52% of candidates reported that receiving too little information during a recruitment process was the greatest source of negativity[1]. ChatGPT can add value if a company employs individuals still getting up to speed with their market and lack market knowledge.
3. Equal Opportunity
Perhaps the biggest advantage that AI systems bring, in an era where equity and diversity are top of many company agendas, is that they can be programmed to avoid unconscious biases in recruitment processes by bypassing the primary sources of bias – names, gender, religion and age[1]. This is deemed to be so important that three in eight professionals believe that solving unconscious bias is the core purpose of introducing AI into recruitment[2]. Chatbots can also address diversity challenges by identifying gendered or biased language used in job specifications[3]. This helps create a welcoming environment for all backgrounds and identities. Without AI, pinpointing exclusive language can be challenging[4].
What are the potential drawbacks?
The advantages are clear, but the route down the AI rabbit hole does introduce two main problems: it reduces human contact, while potentially exacerbating prejudice if incorrectly applied. Why would a firm use Dryden Search if AI can find potential profiles?
1. Lack of human interaction
While technology can provide a better candidate experience, it can also be impersonal[1]. The root problem with any automated recruitment is the lack of human interaction[2], which explains why 82% of people indicated that they wanted to know if they are not going to be talking to a real person during a hiring process[3].
Technology cannot replicate the delicate conversations that recruiters have behind the scenes. Conversations linking directly to company culture, personal values, ambition objectives. It is the personal touch, intuition and relatable experiences that often gets a deal over the line. A chatbot cannot predict this. Final interviews should never be handed over to chatbots because candidates cannot effectively choose whether to join a company without having a human interaction[4].
2. Bias
While AI can offer advantages in reducing bias, it is also true that there are circumstances where it could increase it, because just as AI software can imitate human interaction, it can also imitate human prejudices[5]. If bias is written into an algorithm, it will result in discriminatory hiring practices which might be greater than they would be without its use. The famous example of this is Amazon’s AI-base evaluation model, which was trained incorrectly and therefore was shown to discriminate against female candidates by automatically assigning them lower scores. The tool looked back over the previous 10 years of hiring to determine the characteristics of high performers at Amazon, but because the majority of these benchmark candidates were men, the AI system learned to discriminate against resumés with female designations[6]. Microsoft’s AI tool Face++ – used to analyse applicants’ emotions and personality characteristic – had a similar bias issue whether it automatically assigned black men with more negative qualities than white men[7]. So, although AI can eliminate hiring bias, there is a risk human biases are coded into AI models they create.
Where does it go from here?
Whether recruiters view AI as a threat or opportunity will probably depend on which tasks, they perform within their firms, and whether management introduce AI as a complement or a substitute[8]. It is undoubtable that AI will continue to be woven into aspects of recruitment, but the drawbacks must be mitigated. Unlike so many other professions, the human interaction required in recruitment is invaluable. If you call Dryden we’re happy to say you will speak to a person vs a chatbot and that’s unlikely to ever change!
References:
Black, J.S and van Esch, P. (2020) ‘AI-enabled recruiting: What is it and how should a manager use it? Business Horizons, Volume 63, pp. 215-226. Available at: https://doi.org/10.1016/j.bushor.2019.12.001
Dennison, K (2022) ‘Are AI Recruitment Tools Ethical And Efficient? The Pros And Cons of ATS,’ Forbes Available at: https://www.forbes.com/sites/karadennison/2022/06/27/are-ai-recruitment-tools-ethical-and-efficient-the-pros-and-cons-of-ats/?sh=20e3cb52e4ff [Accessed 18/05/23]
Doyle, O (2023) ‘How to use ChatGPT in Recritment [12 sample use cases]’ occupop Available at: https://www.occupop.com/blog/how-to-use-chatgpt-in-recruitment-10-sample-use-cases [Accessed 17/05/23]
Johansson, J and Herranen, S (2019) ‘The application of Artifiical Intelligence (AI) in Human Resource Management: Current state of AI and its impact on the traditional recruitment process,’ Jonkoping University Available at: https://www.diva-portal.org/smash/get/diva2:1322478/FULLTEXT01.pdf [Accessed 18/05/23]
Mujtaba, D.F and Mahapatra, N.R (2019) ‘Ethical Considerations in AI-based Recruitment,’ Society on Social Implications of Technology, pp. 1-7. Available at: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8937920&tag=1 [Accessed 18/05/23]
[1] RFS (2023)
[2] Dennison, K (2022)
[3] Black, J.S and van Esch, P. (2020)
[4] Black, J.S and van Esch, P. (2020)
[5] Doyle, O (2023)
[6] Black, J.S and van Esch, P. (2020)
[7] Dennison, K (2022)
[8] Black, J.S and van Esch, P. (2020)
[1] Johansson, J and Herranen, S (2019)
[2] Johansson, J and Herranen, S(2019)
[3] RFS (2023)
[4] Recruiter.com (2023
[1] Black, J.S and van Esch, P. (2020)
[1] Johansson, J and Herranen, S (2019)
[2] RFS (20230
[3] RFS (2023)
[4] Johansson, J and Herranen, S (2019)
[5] RFS (2023)
[1] RFS (2023)
[2] Mujtaba, D.F and Mahapara, N.R. (2019)
0 Comments
Leave a comment