AI recruitment tools may be blocking mid-career women from returning to work

Stacey Duguid had spent decades climbing the corporate ladder in fashion, moving through senior roles and periods of freelance work. At 52, she decided to pursue what she called her "one last job." What followed was sixteen months of sending applications into what felt like a void. The automated rejection emails trickled in, but meaningful responses barely came at all. For someone who had never struggled to find work before, the silence was disorienting. She began to wonder if the problem was her.

When she posted about her experience on social media, asking whether other women her age were facing the same wall, the response was staggering. Thousands of women wrote back with nearly identical stories. Duguid, who lives in north London, began to suspect that artificial intelligence screening tools used by employers might be the culprit. She made a calculated decision: she "Botoxed" her CV, stripping away references to her age and the full span of her experience. "AI holds a mirror up to society," she said. "It's a reflection of our bias."

The BBC spoke with more than sixty women between the ages of 40 and 65 across multiple industries, and their accounts painted a consistent picture of rejection at scale. Koeyli Jaluka, 49, had spent two decades in senior positions across manufacturing, corporate, and healthcare. For the previous decade, she had been headhunted regularly. Nine months after being made redundant, she had applied for 442 jobs. Only a handful had responded. When she was told she was "too senior," she understood the coded language. Anna Cowie, 52, had built a thirty-year career in advertising, most recently helping to launch the new Battersea Power Station for major brands. After nearly four years of unemployment, she was now collecting jobseeker's allowance. Both women described the psychological toll: confidence eroded, shame accumulated, the sting of being invisible despite their credentials.

The concern extends beyond individual stories. The City of London Women Pivoting to Digital Taskforce, which works within the financial and professional services sector, has raised alarms about AI's role in screening applications. Their research found that AI systems often fail to recognize the skills that experienced women bring to roles. Career gaps caused by childcare responsibilities, for instance, can be flagged negatively by algorithms that do not account for the context or the transferable skills developed during those periods. The taskforce estimates that without significant retraining investment, AI and automation could displace hundreds of thousands of women's jobs by 2035, potentially costing firms more than £750 million in severance costs alone. Caroline Haines, chair of the taskforce, framed the stakes plainly: "When the country is desperate for economic growth, we need to use whatever resource we have, and the resource of women who are mid-career, mid-experience. If they drop out of the market there are huge repercussions."

The mechanics of the problem are opaque by design. Laura Holden, an AI lawyer and founder of the Bonsai AI and legal consultancy, explained that companies often do not understand how the tools they deploy actually work. CV screening systems grade applications from A to D, encouraging recruiters to prioritize higher scores without understanding what factors drive those rankings. Some tools flag a seven-year career gap as something to note, potentially poisoning a hiring manager's perception before they have even read the application. Holden pointed out that while vendors market these systems as decision-support tools, they often function as automated decision-making systems that reject large numbers of applicants based on opaque algorithmic scores. She called for mandatory testing and stronger regulation before deployment, noting that existing laws were not designed with AI in mind.

Dr. Eleanor Drage, a senior researcher at the University of Cambridge, argued that the concerns are "very justified." Human recruiters, she noted, are better equipped to recognize the value of candidates returning from career breaks and the skills they have developed elsewhere. AI systems, by contrast, operate on logic that is "completely faulty" because they cannot assess a person's intrinsic value or personality the way a human could. Many recruiters do not fully understand the tools they are using, and Drage suggested that companies should use AI to evaluate their own hiring practices rather than to screen "vulnerable populations, like older women going back into the workforce, or people who've had a career break."

Workday, a major provider of AI-powered human resources software, is currently facing a lawsuit claiming it violated California law by weeding out job applicants for discriminatory reasons. The company denies the allegations, stating that its tools do not make hiring decisions but rather help recruiters identify candidates who match job descriptions. It emphasizes that its AI is "rigorously tested" and does not consider age or gender. The government has acknowledged that while AI can help employers manage application volume, recruitment processes must remain fair and transparent, and existing equality and data protection laws already apply to these systems.

Meanwhile, Stacey Duguid has channeled her experience into action. She has set up a community platform to help women feel less alone and to push companies to address the structural problems at play. "I want us, as a group, to be heard and seen," she said. "And for gendered ageism to become a thing of the past. Because if we're lucky, we all get old."