In an age when artificial intelligence (AI) tools can write CVs and cover letters, sort through job applications and onboard new employees, what does this mean for the traditional, human-centric, recruitment market? Can a machine really judge who is the best candidate? Can the system be manipulated? And, what potential is there for AI in recruitment?
To give us a local perspective on how this technology is being used in Bermuda, we spoke to Marisa Whitecross and Liz Ebbs-Brewer, both recruitment services managers at Expertise Group, who pointed out that firstly, local companies don’t tend to suffer from too many applications.
“Investing in AI recruitment tools makes financial sense for large, global firms that manage thousands of applicants per month. In Bermuda, most organisations are small and typically receive a manageable number of applications, so traditional HR practices remain efficient and cost-effective,” Ms Whitecross said.
Ms Ebbs-Brewer also pointed out that Bermuda’s companies sometimes have the opposite problem: “The challenge for employers is not that we have 600-plus applications to review, it’s that we sometimes receive only a handful. Consequently, even candidates who are a weak match to the job requirements get looked at closely.”
This doesn’t mean that there is no beneficial role for AI in Bermuda’s recruitment industry, however. It can be particularly helpful for candidates, when used wisely: “It enables candidates to present their experience in the strongest and most relevant way relative to a specific job’s requirements,” explained Ms Ebbs-Brewer.
“Before AI, many applicants submitted generic CVs and cover letters with very little tailoring, which made it harder for employers to assess whether their background truly matched the role. If used well, AI helps people express their strengths more clearly.”
Honesty and accuracy however, should never be compromised, warned Ms Whitecross: “AI should clarify a person’s experience, not fabricate or embellish. Recruiters can spot that instantly. Employers still want authenticity, not a machine-generated persona.”
There have been reports about candidates manipulating their applications to ensure they get noticed by AI tools. The effectiveness of this, however, is exaggerated: “There are some myths out there about ‘tricking’ AI screening systems – things like including CVs with hidden keywords in white text,” said Ms Ebbs-Brewer. “In practice, serious employers aren’t using tools that can be gamed that way, and HR professionals often easily detect manipulation.”
Such manipulation can also backfire. “Recruiters, as well as AI, respond best to clarity, not gimmicks,” she said. “A simple, well-organised CV that plainly highlights relevant experience will always outperform attempts to game the system.”
Another beneficial way AI could help candidates in the future, is interview preparation. “Interviewing is an infrequent experience for many, so it can be nerve-wracking,” said Ms Ebbs-Brewer. “In the future we foresee an AI interviewer will be able to role play with an applicant and coach them on how they perform.”
AI could also be used to make the administrative process more efficient and even reduce bias, but Ms Ebbs-Brewer has concerns about the latter. “On the one hand, AI tools can perhaps reduce bias, for example, cleansing CVs of data which could create unconscious bias for humans in the hiring process, before a hiring manager initially reviews them.” But, on the flip side, she added: “AI actually making a choice to exclude an applicant might be biased due to the data the AI was trained on.”
Ultimately, Ms Whitecross believes that AI will not fundamentally change recruitment: “As long as humans are doing the work, humans will want to select humans. Being sentient, humans are more capable than AI to judge who is best suited to work with them.”
