If you’re looking for work, there’s an increasing chance that the job description you’re reading was written by artificial intelligence.
Is it good or bad to have AI generate a job description? Human resources experts are cautiously optimistic about the possibilities and the outcomes. Canada’s Human Resources Professionals Association (HRPA) notes that generative AI tools such as ChatGPT can be trained to write effective job descriptions.
But while generative AI may help to automate certain day-to-day HR tasks, “there are still many HR duties AI can’t replace,” HRPA says.
The use of AI by Canadian companies appears to be growing fast. In April 2023, HRPA reported that two thirds of Canadian HR departments had no plans to use AI, and only 25 per cent of companies were using or planning to use the technology at all.
However, a new survey by KPMG in Canada finds that among 872 Canadian organizations, 61 per cent of respondents have already starting using generative AI across various departments, and among those that have not yet done so, nearly eight in 10 (78 per cent) plan to implement it within the next 12 months.
“I know that many organizations are using AI successfully to write job descriptions, but I think there still needs to be a balance,” Lewis Curley, a partner at KPMG Canada’s People and Change Practice in Toronto, says in an interview.
In a research paper it released in September, KPMG notes that AI is reshaping the recruitment process in two ways – by improving job descriptions and allowing for increased candidate personalization.
“The balance is between efficiency through automation and being able to represent the culture and uniqueness of the organization that’s posting the job. You want the job description to be written so the right person will want to apply for the work,” Mr. Curley explains.
“At my company, our recruiters might use a program like ChatGPT to create the outline of a job description, but then they’ll revise it,” says Kathryn Tremblay, co-founder and owner of Altis Recruitment, with offices in Toronto, Ottawa and Vancouver.
“Human recruiters review the description to make sure it’s inclusive and free of any bias and add language about the company’s culture and values, along with any interesting job responsibilities, perks or benefits. It’s also important that each description reflects the company’s brand voice because, since it’s the candidate’s first connection with their next potential employer, you want to make sure the description is authentic,” she says.
“Like every other kind of tool we bring into the digital workplace, AI can be helpful to human resources because it can handle so much work that’s repetitive, and that includes writing job descriptions,” says Pamela Lirio, associate professor of human resources management at the University of Montreal.
“But at some point a human being needs to put their voice to the job description. Otherwise, tools like ChatGPT will just come up with boilerplate wording that amalgamates information it finds on the internet,” she says.
More and more HR professionals are turning to AI-based applicant tracking systems (ATS) to help organize their hiring and personnel management processes, including preparing job descriptions, Prof. Lirio says.
Companies can buy ATS tools off the shelf or create their own, which is what financial companies like to do for security reasons. An AI-based ATS helps companies sift through resumés faster, focusing on those that have words and phrases that come closest to matching the job description, which may also have been AI-generated, she explains.
“It’s practical because with digitization and AI, there are so many more resumés that arrive to companies and recruiters, and job applicants have access to so many more postings than they did in a pre-digital world,” Prof. Lirio says.
Recruiters and applicants alike should keep in mind that AI is still evolving and is far from infallible, says Shaunna-Marie Kerr, the associate director of career advising and design at the Centre for Career Design at York University’s Schulich School of Business in Toronto.
“There’s still a lot of training to be done. Some people think of AI as a kind of souped-up spell check, but you have to be aware of the nuances that an AI-generated document can generate,” she says.
Recruiters or applicants can ask AI to generate wording that sounds fantastic, “but you have to make sure you see yourself or your organization in the wording – something that job seekers can match with their skills and experience,” Ms. Kerr says.
“Employers are now receiving up to 15-per-cent more applications than before AI because job seekers are using programs that automate the application process,” Ms. Tremblay says. “This is why I think human connection and conversation are still so critical in the hiring process. The resumé might get the candidate the interview, but the hiring manager vets the experience and skills,” she says.