The future of HR isn’t adopting AI – it’s redesigning work around it

Simply embedding AI into legacy workflows can create friction rather than flow, writes Georgia Russell.

For many HR teams, the first wave of artificial intelligence has felt familiar. New tools arrived promising speed, efficiency, and automation. Job ads were drafted faster, policies summarised in seconds, and reports produced with less effort. The underlying assumption was simple – add AI to existing HR work, and productivity will improve. AHRI research shows that most organisations have focused their early AI use on speeding up existing HR tasks, rather than redesigning the work itself.

That assumption is now being tested. National AI research has found that while AI is increasing efficiency across people functions, it does not consistently improve decision quality or outcomes without changes to how work is designed. What many organisations are discovering is that while AI can accelerate work, it does not automatically create value. Faster outputs do not equal better outcomes. In some cases, AI has simply amplified inefficiencies that were already there. The real shift now underway is more fundamental. HR is moving from trying to adopt AI to redesigning work around it.

Why layering AI onto old models falls short

Most HR operating models were designed for a world where information was held tightly by a few, data was fragmented, processes were manual, and human effort was the primary constraint. Structures, hand‑offs and governance evolved to manage that reality. Governance bodies have observed that many operating models were built for environments where information was scarce and manually processed, not instant and automated. AI changes the constraints.

When information is instant and more accessible, and analysis is automated, the bottlenecks move. Decisions, judgement, prioritisation and trust become the limiting factors. Simply embedding AI into legacy workflows can create friction rather than flow. Australian regulators have warned that AI can amplify existing process weaknesses and bias when layered onto legacy models without redesign.

For example, automating parts of recruitment without rethinking role design, process, or decision rights can speed up hiring activity while increasing rework or bias. Or, using AI to generate people insights without changing how leaders engage with data can lead to more dashboards, more confusion, and fewer decisions. Optimisation has a ceiling. Redesign does not.

Redesigning HR work starts with the work itself

Redesigning HR work around AI requires stepping back from roles and focusing on both the higher-order outcomes and then jumping into the detail – the tasks. Which activities truly require human judgement, empathy or context? CSIRO’s work on responsible AI stresses the importance of deliberately deciding which people‑related decisions must retain human oversight and judgement. Which can be reliably supported or handled by AI? Are some tasks obsolete? Where does oversight need to sit to ensure quality and accountability? This approach shifts HR away from job descriptions built around activity and towards work designed around outcomes and accountabilities. It also reframes AI as a collaborator rather than a replacement.

In practice, AI increasingly does the first pass. It drafts, analyses, surfaces patterns and flags risks. Humans step in to interpret, challenge, decide, and act. The value of HR work moves upstream. Australian policy research indicates that AI is increasingly performing the first pass of analysis and synthesis, shifting human value towards interpretation, challenge and decision making. The real challenge is letting go of work that no longer needs to be done by humans, and being willing to redesign processes that feel familiar but are no longer fit for purpose.

What this means for HR roles

As work is redesigned, HR roles start to change shape. Transactional work continues to reduce, not because it disappears, but because it is handled differently. Employee queries are resolved through intelligent self‑service. Data flows continuously rather than being manually compiled. Reporting becomes predictive rather than retrospective. Research, reporting, and/or survey results insights, thematics, and analysis are almost instantaneous.

At the same time, advisory and specialist roles become more critical. HR practitioners spend less time producing information and more time making sense of it. The Australian Human Rights Commission notes that AI‑enabled HR work places greater emphasis on critical thinking, ethical judgement and influence, capabilities that cannot be automated. Partnering with business leaders to navigate complexity, trade‑offs, and change. This shift also places greater emphasis on skills like critical thinking, ethical judgement, influence, and systems thinking. These capabilities cannot be automated, but they can be amplified when the right work is taken off human desks.

Operating models need to evolve

Redesigning work around AI also challenges how HR is organised. Traditional models that separate business partnering, centres of expertise and shared services were built to drive efficiency and consistency. AI reduces many of the constraints those models were designed to manage. Australian government research suggests that rigid separations between service delivery, expertise, and partnering are increasingly misaligned with AI‑enabled ways of working. Without redesign, AI simply accelerates the old model. With redesign, it enables a different one.

More flexible, outcome‑based ways of working are emerging. Cross‑functional teams form specific workforce challenges. Data and insights are shared rather than owned. Decision making is pushed closer to the work. This does not mean abandoning governance. In fact, governance becomes more important as AI takes on a greater role – it just might take on a different “look and feel”. Clear accountability, escalation pathways and human oversight are essential to maintain trust and compliance.

HR’s moment to lead

This shift creates a defining moment for HR. Redesigning work around AI is not a technology project. It is an organisational one. It affects how work is done, how decisions are made, and how value is created. These are areas where HR has deep expertise and influence. If HR steps into this space, it can become a central architect of the future of work. If it stays focused only on adoption, it risks being a passive user of systems designed by others.

The move from adopting AI to redesigning HR work around it is not a trend. It is a structural change. HR leaders who recognise this will shape the function’s relevance for the next decade. Those who do not, may find they have optimised themselves out of influence. This kind of redesign work is hard to prioritise amid the day‑to‑day demands of BAU, but the pace of change means it can no longer be deferred; the time to do this work is now, and it is critical.

Georgia Russell is the executive director of consulting at shilo.

Want to see more stories from trusted news sources?
Make HR Leader a preferred news source on Google.