.png)
If the news and social media chatter is anything to go by, artificial intelligence will eventually take over many of our work tasks – including how we recruit and hire people.
AI does have huge potential to make hiring managers’ and recruiters’ lives easier, and plenty of employers are already using it in some form, whether it’s to write job ads or help filter hundreds of applications for a single role. According to an HRNZ survey, 27% of HR professionals use AI for tasks such as CV screening, with some respondents saying they also use it to source candidates.
As more organisations get tempted by the convenience that AI recruiting tools may offer, it’s important to look at the pitfalls and risks they can create relating to bias, which have real implications for diversity and inclusion in the workplace.
How is AI Used in Recruitment?
There is already an abundance of software tools that use AI to analyse resumes, create shortlists of candidates and even perform video interviews.
The most common examples of AI recruiting tools are applicant tracking systems (ATS), which can use algorithms to filter incoming CVs based on specific keywords related to skills, qualifications and work history. In recent years (especially after COVID), AI video interviewing tools have also become popular, with some products claiming the ability to identify a candidate’s personality traits and score them on suitability for the role. Other AI-based tools can analyse applicants' social media profiles or even potential candidates to make recommendations about their potential fit.
What the Research Says About AI Recruiting
A growing body of research indicates that AI recruiting tools can unintentionally introduce more bias into the hiring process or reinforce existing biases related to gender, age, ethnicity and other characteristics.
The quality of data is a key issue in AI bias. If an AI tool relies on data that is already biased, it will likely reproduce that bias. As one researcher said, ‘AI might learn, based on the data, that people called “Mark” do better than people named “Mary” and are thus ranked higher. Existing biases in society are reflected in and amplified through data’.
According to one experiment by Bloomberg, when ChatGPT 3.5 was asked to rank resumes, it favoured candidates with names from specific demographics to the extent that its suggestions would fail to meet anti-discrimination benchmarks if used in a real-world setting.
In another experiment, researchers created an AI model to simulate job advertising, where job ads are scored based on how well they attract candidates and then improved over multiple rounds. This led to one group of candidates being consistently attracted at a greater rate despite the scoring system being designed not to consider bias at all. They also found that ‘biases in data can be made worse by the human-created algorithms supporting AI, which often contain human biases in their design’.
Mitigating Bias in AI Recruitment
The best way to mitigate bias around artificial intelligence and recruitment is, quite simply, human intervention. Hiring managers, recruitment, and HR teams remain the decision-makers when it comes to hiring, so their oversight is vital.
Having recruiters and hiring managers conduct in-depth interviews and reference checks is essential for identifying suitable candidates and assessing them based on their actual abilities and not their identity.
Oversight could also include establishing clear organisational policies for artificial intelligence recruitment, regular audits of hiring processes and decisions that involve AI tools, and ensuring the data used to train AI tools is diverse and truly representative of the people of New Zealand.
Using Artificial Intelligence Recruitment Tools for Good
Despite AI’s risks and limitations, we can expect that organisations will continue to use it and incorporate it into even more aspects of their recruitment process. Although human involvement in recruitment cannot be fully replaced by AI, perhaps the best way to use AI is to reduce time spent on administrative tasks that do not require decision-making or judgment calls.
Further, AI can already be put to good use in supporting diversity and inclusion in hiring measures. AI tools can be used to remove names and demographic information from CVs and job applications and reduce hiring managers’ unconscious bias. They can also be used to add more inclusive language to job descriptions and job ads.
The Future of New Zealand Recruitment Needs Humans
The human touch is the core component of what we do here at Beyond Recruitment. Regardless of what new technologies might appear as the next ‘big thing’ in recruitment, we will continue to use our decades of collective experience and our values of diversity and inclusion to help businesses discover exceptional hidden talent.
Our deep industry knowledge allows us to refine recruitment strategies in line with the nuances of an organisation’s culture, role requirements and overall vision. Get in touch with us today to discover how we can help you find people the algorithms won’t.
