At a time when artificial intelligence (AI) seems to be infiltrating every element of our lives, it is not surprising that HR departments and agencies are looking towards the capabilities of AI to streamline processes and even strengthen workforce recruitment and development.
AI for HR promises to call upon information that is carefully curated to produce complex algorithms which work alongside machine learning to provide a more efficient, unbiased recruitment process. The result? Time and money saved for organizations and a streamlined, fair and efficient recruitment process for applicants. If this sounds too good to be true, that could be because it is. The use of AI and by association, the reduced culpability of human workforce, brings with it a host of ethical considerations. The technology jobs available on Motion Recruitment show the range of tech opportunities that are available currently. This wealth of opportunities is in part due to the huge rise in demand for tech jobs, as well as the rise in remote working, which means that new talent is not restricted to a specific geographical area.
There may well be a place for AI in HR, but it is more important than ever that HR departments and agencies work hard to achieve the fine balance between efficiency and improved user experience through innovation, and the ethics of allowing technology to determine the course of someone’s life. As outlined at the HR & Technology Conference and Expo in Vegas, the use of AI in even its earliest, most seemingly insignificant form, must come with an awareness of potential ethical implications.
Addressing unconscious bias
For leaders that assume that the use of AI will remove unconscious bias, it may be a good idea to reevaluate. In fact, biased decision-making is one of the central ethical concerns surrounding AI-led HR. AI algorithms develop from historic data and provided information. Therefore it stands to reason that AI will identify historical trends, including unconscious or conscious bias, and weave that into the algorithms. This can perpetuate deep-rooted biases, which may not be easy to identify. A real-world example of this is Amazon’s AI recruitment process as highlighted by Forbes, which was scrapped in 2018. The AI tool was utilized to shortlist applications based on shortlisted resumes submitted over the past decade. That sounds great in theory, but in practice, most of those shortlisted had been men. As a result, the AI tool applied the gender bias as a “rule” and continued to apply it for new applicants. The danger comes because by nature, unconscious bias is not something that we are aware of, therefore we are unable to mitigate against it in the first instance and notice it during the recruitment process. After all, if most interviewees have been men, it is unlikely that a hirer is likely to comment on the status quo remaining.
Transparency
AI algorithms can be complex and opaque, making it challenging for individuals to understand how decisions are made, or identify biases. Ethical AI-driven HR systems should prioritize transparency, providing explanations for the reasoning behind decisions. Moreover, organizations must establish mechanisms for accountability in case of algorithmic errors or biases, ensuring that human oversight and intervention are available when needed.
Let’s return to the Amazon example. If you are aware of the bias, you can mitigate it by removing gender specific data, however there are more subtle nuances (such as part time working, maternity leave breaks, or even schools and colleges attended/ education history) that are a subtle indicator of gender and much harder to isolate in a series of complex data designed to sift through high volumes of resumes.
Privacy, data protection, and autonomy
AI-driven HR systems require substantial amounts of personal data to function effectively. Organizations must handle this data with the utmost care to respect individuals’ privacy rights. Adequate data protection measures, informed consent, and clear communication about data usage are crucial to build trust between employees and AI-powered HR solutions. The data protection requirements for AI-powered recruitment may far exceed those for manual, person-led processes; reviewing and appropriate removal of data must therefore be factored into the process.
Autonomy is another factor for consideration. In AI-led candidate assessment applicants must be made aware AI’s role in the decision-making process. Candidates have the right to understand how their data is being used and how AI influences hiring choices. Organizations should ensure that the use of AI does not compromise the autonomy of individuals, and they should be given the option to request human review of decisions.
Complementing – not competing – with HR professionals
Resisting AI may well be futile; at best it may mean delaying the inevitable. For most organizations, it may be more beneficial to embrace change and, in doing so, understand the opportunities and limitations of AI in HR processes. AI models shouldn’t be allowed to or expected to displace HR professionals or perpetuate ingrained bias. With experienced HR teams at the helm, it is possible to ensure that programmes enhance HR practices while respecting the rights and well-being of employees by fostering a workplace where technology and humanity coexist harmoniously.
