a day ago
- Business
- Business News Wales
AI Tools Could Help to Address Recruitment Bias, Say HR Specialists
AI has long been used to streamline high-volume recruitment processes, but HR leaders say its potential to help reduce bias in hiring decisions is now coming into sharper focus.
While questions remain over how AI models are trained and applied, employers including Network Rail and specialists at DeepLearnHS believe carefully implemented tools can help challenge human assumptions and promote more consistent, skills-based decisions.
Colin Minto, Talent Acquisition Leader at Network Rail, said his organisation is using inclusive language tools to reduce barriers at the very start of the recruitment journey.
'One of the first things you do when you're recruiting is advertise roles,' he said. 'If you are writing those roles, and a number of people write adverts across the organisation, [you need to] make sure you're not using gendered or potentially exclusionary language. You can have tools that pick that up and recommend more neutral alternatives. That helps attract a more diverse pool of candidates.'
Chris Butt, CEO and founder of DeepLearnHS, agreed that human bias often enters the process unconsciously, and that AI – used well – can offer a helpful counterbalance.
'Human beings have been conditioned from the moment they're born,' he said. 'It's very difficult to work against that bias. But when you're building a model, you can start to remove some of the biases we understand – things like gender, age, and ethnicity – from the data being processed.'
Chris added that using AI to spot patterns in large datasets could also help identify early career pathways, enabling fairer access to opportunity from a much younger age.
'Traditionally, you'd assume someone finishes school and then goes into work, university or training,' he said. 'But you can do so much more to understand what that person looks like throughout their schooling, using data, so we can fine-tune the best route forward. AI can help with that stitching process.'
Mark Kunaseelan, Head of Resourcing at University of the Arts London, said his team has yet to adopt AI widely, but is cautiously exploring future applications. One of his hopes is that AI might offer greater objectivity during shortlisting, as long as final decisions remain with people.
'In my head, I think of it as: AI is going to tell you where to begin your shortlisting,' he said. 'Here's a ranking – but we still need the human element, to double check that's correct. Sometimes I've seen scoring that's completely off from how I would've scored it, but then I think, am I flawed? Do I have biases? Affinity bias, or whatever it is?'
All three leaders stressed the need for transparency in how AI systems are designed and implemented, with clear governance over data use and decision-making. While no tool can fully remove subjectivity from recruitment, there is growing recognition that – used responsibly – AI may help challenge some of the bias that has long gone unexamined in human-led hiring.
Listen to Chris, Mark and Colin discuss this and more in episode one of the DeeplearnHS podcast series, AI in Hiring & Workforce Strategy, here
