According to Barb Hyman, founder and CEO of London-based recruitment platform Sapia.ai, resumes are “broken.”
“How many people are we missing out on because we rely on artifacts?” Hyman added that AI can now tailor resumes (known as CVs in the UK) to fit any role.
Graduate employers in the UK will receive an average of 140 applications for every vacancy in 2025, the highest level in 30 years, according to data from the Student Employers Association. In a crowded marketplace where candidates can create endless resumes and cover letters using AI, finding the right talent can be more difficult than ever.
West End 61 | West End 61 | Getty Images
Hyman said skills, not work history, determine who gets hired. For example, a cosmetologist may not seem like an obvious fit for a contact center job on paper, but cosmetologists may be well-suited for the role because they are highly empathetic, tend to listen to people all day long, and spend hours solving complex problems, she told CNBC.
According to Hyman, a direct assessment of skills can reveal strengths and weaknesses hidden in a traditional resume, and can reveal situations that even the candidate wasn’t sure about.
Skill-based hiring is on the rise
Hyman isn’t the only one who thinks hiring is moving in this direction. Around two-thirds of UK organizations use skills-based recruitment, according to a 2026 study by Michael Page, a recruitment firm headquartered in Surrey, UK. Almost 60% of candidates say they are more likely to apply for a role defined by their skills rather than traditional qualifications or background.
This is causing employers to think carefully about which skills they actually value. James Broadhead, Dublin-based CTO of AI decision-making company Taktile, told CNBC that thanks to AI, prototypes that once might have taken engineers three weeks to create can now be produced in one afternoon.
You don’t know if someone has the skills to do something until you try.
“The hard part is having the judgment to know which ones are good,” Broadhead said.
He added that Taktile is now looking for interpersonal skills in candidates, such as how well they can talk to customers, as well as a willingness to be self-reflective and question what they’ve built. They’re also interested in whether candidates are experimenting with AI on an ongoing basis.
“The pace of model releases is so fast that the upper limits of what you can do change every three months,” Broadhead said. “So what we’re asking is: What have you changed about your AI tools in the last three weeks?”
Other companies evaluate hiring candidates by putting them on the bottom line. At digital marketing and advertising agency PMG, graduate candidates are expected to complete practical exercises individually, work in teams, and work on client briefs. Josh Krzyzewski, president of EMEA at London-based PMG, told CNBC that this is meant to assess how candidates actually engage and approach the job.
Rudy Suardi | E+ | Getty Images
“You don’t know if someone has the skills to do something until you try,” Krzyzewski said. “The past doesn’t tell us that.” Younger candidates are often better suited for the field because Gen Zers have fewer preconceived notions of what the job looks like, he said.
“This is another reason why investing in young people is so important to us, especially at a time when the world is changing at breakneck speed,” Krichevsky said. “Graduates don’t come in with fixed assumptions about how work has been done.”
Breaking away from human prejudice
Perhaps paradoxically, even though AI is exacerbating some of the traditional employment problems, it can also be used to find solutions.
For example, this technology can be used to highlight candidates from diverse backgrounds. Sapia.ai’s Hyman argues that resumes and traditional interviews are always susceptible to human biases, such as “mirror hiring,” where hiring managers consciously or unconsciously favor people with similar backgrounds.
What is broken is human prejudice, but we all deny it.
“We look at things through the prism of our own experiences,” she said. “What’s really broken is human prejudice, but we all deny it.”
For Hyman, that doesn’t mean eliminating human employment entirely. Although the final decision should be left to the new hire and the people working with them, technology is well-suited to screening candidates at scale. Evaluation should always be a two-way street, she said.
“Hiring the wrong person can be extremely painful for recruiters, and they are in the best position to know if you are the right fit for them and their team,” she says. “Job candidates deserve as much opportunity to stress test candidates as you do.”
