How does Rest Less make money
We make money through advertising and commission from affiliate links, which enable us to offer Rest Less as a free service to our users. The content on this page may use affiliate links, which track traffic from our website to a third party provider and enable us to receive a commission or payment from any traffic we refer.
* Affiliate links on this page have an * next to them. We place enormous importance on our editorial independence and the integrity of our content which means that we will never change how we write about something as a result of an affiliate link.
From the way we browse the internet to how we choose our next Netflix show, AI is rapidly changing many aspects of our lives – and the same is true for the world of recruitment.
Companies are now leveraging a range of AI-powered hiring tools to automate their recruitment processes, reach a wider talent pool, and eliminate human bias. But are these tools actually free from prejudices like ageism?
The problem with AI hiring tools
HR News reports that, as of 2024, 48% of UK recruitment agencies have embraced some form of AI technology to find, interview, and hire candidates.
Tools range from CV scanners to software that analyses applicants’ facial expressions, speech patterns, and body language during video interviews. And while they’re ostensibly designed to help companies find better quality talent and eliminate prejudice, this isn’t always the outcome.
As journalist Hilke Schellmann, author of The Algorithm: How AI Can Hijack Your Career and Steal Your Future, tells the BBC, “We haven’t seen a whole lot of evidence that there’s no bias here […] or that the tool picks out the most qualified candidates.”
For example, back in 2023, the U.S. Employment Opportunity Commission (EEOC) settled its first lawsuit regarding AI bias in hiring with China-based tutoring firm iTutorGroup Inc. According to JD Spura, a leading source of legal intelligence, “evidence suggests that the tutoring company programmed its recruitment software to reject older applicants (based on birthdate entered in the application) in violation of the Age Discrimination in Employment Act (ADEA).”
Suspicions of ageism were raised when a rejected candidate resubmitted an identical application with an altered birthdate to make themselves appear younger, only to be offered an interview. Overall, authorities allege that, in early 2020, iTutorGroup failed to hire more than 200 qualified older applicants because of its biased AI screening tool(s).
A lack of awareness and accountability
In the case of iTutorGroup, the company seems to have intentionally deployed an AI hiring solution to screen out older applicants. However, some companies may be unintentionally using biased AI-powered hiring tools with discriminatory results.
Often, these tools are essentially ‘black boxes’, say researchers from Cambridge University, meaning their internal workings aren’t revealed to users. The Cambridge researchers also say that there’s “little accountability for how these products are built or tested […] As such, this technology, and the way it is marketed, could end up as dangerous sources of misinformation about how recruitment can be ‘de-biased’ and made fairer.” Even the vendors may not know exactly how their tools are working.
Schellmann has similar concerns, telling the BBC she’s worried that developers are hurrying along underdeveloped or faulty AI hiring tools to capitalise on high demand. “Vendors are not going to come out publicly and say our tool didn’t work, or it was harmful to people”, she explains, while organisations that have already used them may be “afraid that there’s going to be a gigantic class action lawsuit against them”.
Why aren’t AI-powered hiring tools unbiased?
The problem with AI-powered tools is that they’re trained on past data, which likely contains human biases. As the Discussion Paper published for the AI Safety Summit organised by the UK government says: “These biases, often subtle and deeply embedded, compromise the equitable and ethical use of AI systems, making it difficult for AI to improve fairness in decisions.”
To use a simplified example, say a company displays a pattern of biased hiring practices (whether intentional or not). This bias will become embedded in any AI algorithms that are trained on their hiring data. Therefore, when implemented, these tools will display similar discriminatory behaviour, resulting in conformity rather than diversity.
And simply removing attributes like age, race, and gender isn’t enough to remedy this issue. As the AI Safety Summit Discussion Paper puts it, “models can infer these attributes from other information such as names, locations, and other seemingly unrelated factors.”
Take this study of three state-of-the-art large language models (LLMs) from the University of Washington, for example. The researchers found that these AI tools favoured white-associated names 85% of the time, female-associated names just 11% of the time, and never favoured Black male-associated names over white male-associated names.
These associations can be even more obscure. In her book, Schellmann also uses an example of a CV screener that prioritised people who listed baseball as a hobby over people who listed softball. The former is more commonly played by men, and the latter by women.
Final thoughts…
Biases aren’t the only reasons to be wary of AI hiring tools. Simple tests have revealed other issues. For example, Schellmann reports registering a decent score of 73% in a one-way, AI-powered interview by simply reading aloud from a Wikipedia entry in German.
However, concerns over ageism and other forms of discrimination should be enough to make us think twice about implementing AI-powered hiring tools, at least for now. Instead, employers and recruiters looking to become more age-inclusive may be better off re-evaluating their existing, human-driven hiring practices.
You can find plenty more information on age-inclusive hiring practices on our for employers page. Or why not email us at [email protected]?
Has your company taken steps to reduce age bias in your recruitment strategies? If so, do you have any advice for other organisations? We’d love to hear from you in the comments below.
Join the discussion
Read our full commenting terms and guidelines