The study, published in Information Technology & People, draws on hour-long interviews with 41 participants: technology experts, HR professionals, AI companies and advisers operating in the UK, Europe, the US, Australia and the United Arab Emirates.
Professor Elisabeth Kelan found two recurring positions among interviewees. The first — techno-optimism — held that AI could actually reduce the biases already embedded in human hiring. Participants pointed to more representative training data, removal of personal identifiers, and algorithmic auditing as tools for fairer outcomes.
The second position, which Kelan terms techno-hesitation, emerged when participants acknowledged the limits of that optimism. But the telling detail is what happened next: concern about discrimination rarely became a reason to stop. Instead, it became a reason to manage. Ethical questions were repackaged as legal exposure, brand risk, or operational variables an organisation could control.
"What emerged was a form of techno-hesitation, where ethical concerns were reframed as legal, reputational or managerial risks that organisations could manage," said Professor Elisabeth Kelan, a Professor of Leadership and Organisation at King's Business School.
The research does not test whether any specific AI hiring tool is accurate or discriminatory. Its focus is narrower and, arguably, more unsettling: how organisations talk about bias as they continue to adopt the technology. The pattern the study describes — where caution is absorbed into the justification for adoption rather than slowing it — may make AI recruitment appear more manageable than the underlying fairness questions warrant.
For employers, the study's practical recommendation is to start with the hiring problem and the fairness test before selecting a tool. That means being clear about which data and decisions the system will influence, how outcomes will be tested across different demographic groups, who is accountable for decisions, and how applicants can challenge results. For jobseekers, the basic questions are whether AI will screen or score their application, what information it will assess, and whether humans remain meaningfully involved.
The paper, Negotiating algorithmic bias: Discourses of techno-optimism and techno-hesitation in AI-mediated hiring, is published in Information Technology & People.
To stay across the latest in cloud, AI and enterprise tech analysis from Compare the Cloud, subscribe to our weekly newsletter at https://www.comparethecloud.net/newsletter
