AI screening promises to speed up recruitment and cut human bias, but the rush to let algorithms sort applications in minutes looks like it could deepen discrimination and leave many job seekers locked out before a person ever sees their CV.
Right now, more employers are weaving AI into almost every stage of hiring, from scanning CVs to running text and video interviews mainly because they are overwhelmed by volume and want faster, cheaper decisions. In practice this means many applicants now receive rejection emails within hours or even minutes and often suspect a machine has made the call rather than a human recruiter. This shift has grown quietly alongside the broader automation of office work, where HR teams are under pressure to do more with less and rely on software to filter large candidate pools.
Specialist hiring platforms now claim to have run millions of AI‑driven interviews for major brands across airlines, supermarkets and large retailers, promising fairer and more efficient hiring. Their systems usually invite every applicant into an automated chat, explain how the data will be used and ask a small set of open questions about motivation, problem solving and future goals, often with no time limit but strict word guidance. Supporters argue these tools never see age, gender, address or appearance and say that by ignoring personal details and focusing only on written answers they reduce the unconscious bias that can creep into traditional interviews. At the same time, professional bodies note that some employers now use AI to strip identifiable data from CVs before human review, hoping to stop decision makers from judging candidates based on where they live, their background or other demographic clues.
However, researchers and unions are increasingly concerned that handing too much power to opaque algorithms seems to create a new layer of risk for workers, especially those from disadvantaged groups. Studies from Australian universities suggest that the data used to train hiring systems often reflects a narrow slice of the workforce, which can reinforce old patterns of exclusion instead of improving diversity. Legal experts warn that this “algorithm‑facilitated discrimination” is hard to detect or challenge because the models are complex, their decision rules are rarely transparent and rejected candidates usually see only a generic email rather than a clear reason. Academic work in human resource management also indicates that robust organisational policies, legal safeguards and diversity aware HR managers still do more to improve inclusion than AI tools alone, which underlines the need for human oversight in final hiring decisions.
The broader picture is that AI looks set to play a permanent role in recruitment, but without stronger rules and scrutiny it may quietly reshape who even gets a chance to be considered for a job. The federal government has started to respond with a national AI plan, voluntary guardrails and a $30 million AI Safety Institute to monitor emerging risks, yet legal scholars argue that discrimination laws still lag behind and leave gaps in protection for job seekers screened by machines. As employers experiment with ever more automated processes, the balance between speed, fairness and transparency seems to be the real test. Organisations that treat AI as a decision support tool rather than the final judge are more likely to build trust while those that outsource judgment entirely to algorithms risk both reputational damage and future legal challenges.
