Securing a job has often been harder for older people, who may be unfairly dismissed as being less adaptable or creative than younger applicants.
But has the rise of artificial intelligence and its use in recruitment made things more difficult for silver-haired jobseekers to keep a firm footing on the career ladder? A study from Australia suggests this might be the case for those over the age of 45, as advanced technology plays an increasingly influential role in the hiring process.
Alysia Blackham, a professor at Melbourne Law School, part of the University of Melbourne, found that when she asked ChatGPT to recommend people with enthusiasm and new ideas for the tech industry, it suggested early and mid-career professionals, groups that cover people between the ages of 21 to 45. Older people were not suggested.
Prof Blackham also reported that a job advert produced for her by the generative AI tool included phrases such as “a fresh perspective” and “fast-moving”, which may put off older candidates.
Given ChatGPT has been shown to have biases on the basis of gender and race, she told The National it was not surprising that it also appeared to discriminate according to age. This, she said, was probably a reflection of the data it had been trained on.
“We don’t have great information on what they [generative AI programmes] have been trained on, but the best guess is the entirety of the internet, everything that’s out there in the public sphere, and there’s a lot of material on the internet that’s sexist, racist and ageist,” said Prof Blackham, who is the author of books including Extending Working Life for Older Workers.
Responsible use of AI crucial
While there may be nothing new about older job applicants being less favoured for certain roles, Prof Blackham indicated that AI risked taking things to a new level. “If a human recruiter is discriminating, they do it applicant by applicant,” she said. “These tools, for example, when used in recruitment, can reject people nearly instantaneously at scale.”
With technology making it easier to apply for jobs – sometimes all that is needed are a few clicks – it is perhaps unsurprising that employers are turning to AI to screen the deluge of applications.
Mark Daley, chief AI officer at Western University in Canada, said that while AI may be useful for recruitment, it could easily take over the process.
“You think you’re just using the AI for advice and you’re still the one making the decision, but if you’ve got 200 resumes to read, are you really carefully double checking the AI or is it easier to just click yes, approve, approve, approve? This is what we’re grappling with right now," he said. "What do we owe applicants? What do we owe colleagues? What do we owe fellow humans? And where can we bring technology into that in a way that is helpful but doesn’t launder away our agency?”
Prof Blackham said employers should be “extremely cautious” about using AI in recruitment, using tools only after testing them thoroughly first.
"They need to be pushing these technology companies who sell the tools to show them how they’ve considered bias, how they’ve ensured that their datasets for training and testing and are representative, including of older workers, and how they’ve ensured that there is no bias being implemented in these tools,” she said. “Ultimately, it’s an employer who will be held liable if a tool discriminates.”
Addressing bias
Nikhil D’Souza, general manager at HR company Nathan and Nathan in Dubai, said it may be easier to eliminate the biases of AI than of people.
“In some cases, AI may screen out experienced candidates based on indirect indicators such as graduation dates, years of experience, or career patterns. However, human recruiters are not immune to bias either,” he said.
“The key difference is that AI bias can be systematically identified, measured and corrected when organisations implement proper governance, regular audits and ethical recruitment practices.”
Breaking down barriers
Despite ChatGPT’s apparent lack of enthusiasm for older applicants, at least for certain jobs, Prof Blackham is clear that mature candidates can contribute much to the workplace. That includes tech-related roles where some people, probably unfairly, might expect older staff to be less comfortable.
“There are really embedded stereotypes around older workers and how they use technology,” she said. “Partly, these are the result of employers often excluding older workers from training and upskilling. So it’s a bit of a self-fulfilling problem. Often older workers are great at using technology, particularly if they’re given the chance to learn something new.”
Mr D’Souza echoed that point. He said staff who are always keen to learn – and to integrate AI into their work – will thrive, “regardless of age”.
“Older professionals should focus on demonstrating adaptability, continuous learning and digital proficiency,” he said. “Employers today are increasingly seeking a blend of experience, resilience, leadership and the ability to navigate change – qualities that many seasoned professionals bring in abundance.”
Even if AI is not always a friend to older job applicants, it does not mean that more mature candidates looking for jobs, or younger people planning to go to university, should avoid using the technology to optimise their applications.
For a cover letter, or a personal statement for university applications, Prof Daley suggests that people write the initial version themselves before running it through the three main AI programmes – ChatGPT, Claude from Anthropic and Google Gemini – for feedback. Further drafts can go back and forth between the applicant and AI.
He said the jobseeker should bee the one "holding the pen on the last draft, because what you don’t want to do is to submit an application that reads the same as everyone else’s who has used" a tool such as ChatGPT. You want your own original voice in there,” he added.
