RecruitLocal: Can Ethical AI Finally Fix What’s Broken About Recruitment?
By the Upturn People Team · November 2024
The recruitment sector has a problem it has been slow to acknowledge. Traditional hiring processes — from application sifting to interview design — are systematically biased against candidates from disadvantaged backgrounds, underrepresented communities, and those with non-linear career histories.
The results are felt acutely in communities like the ones Upturn serves. Employers complain that they cannot find good local candidates. Individuals complain — with considerable justification — that the hiring process feels designed to exclude them. And communities find themselves trapped in cycles of local unemployment even when local employers have local vacancies.
Where AI Has Failed Before
The application of AI to recruitment has not, historically, made this better. Early AI screening tools — most famously, Amazon’s abandoned recruitment AI — simply encoded the biases of the historical data they were trained on, systematically disadvantaging women, people of colour and candidates from less prestigious educational backgrounds.
The problem was not the technology. It was the assumption that historical hiring decisions were a good proxy for future performance — an assumption that was always dubious and that embedded structural inequality into automated processes at scale.
What Makes RecruitLocal Different
RecruitLocal was developed by Upturn specifically to address these failures. Rather than training our matching algorithms on historical hiring decisions, we build candidate profiles from scratch — mapping skills, values, interests and potential against role requirements, rather than filtering for the proxies (educational credentials, job title history, previous employer prestige) that traditional hiring uses.
Crucially, every candidate in the RecruitLocal network has been supported by Upturn’s People team. We know who they are, what they are capable of and what support they may need to thrive. This human knowledge — combined with algorithmic matching — produces a quality of candidate connection that purely automated systems cannot replicate.
The Results
Employers using RecruitLocal consistently report improvements in candidate quality, workforce diversity and retention. Candidates supported through the platform consistently report that the process felt fair, transparent and respectful — a marked contrast with their experience of conventional recruitment.
We are proud of what RecruitLocal has achieved. But we are also clear that technology is only ever as good as the values and intentions behind it. RecruitLocal works because it is embedded in Upturn’s 20 years of community knowledge — and because it is designed, from the ground up, to serve people as much as employers.
