The future of successful HR will use AI with human oversight

The impact of AI on every business function cannot be understated. But it's impact on HR may be the most profound of all. Every function the HR team undertakes from recruitment and onboarding, through skills development and performance management, to offboarding will be impacted by AI.

HR teams are responsible for managing highly sensitive information and managing processes that directly impact the livelihoods and personal lives of the entire organisation. That means AI initiatives must ensure there is no data leakage and any decisions supported by AI are not carried out in an inscrutable black box. While many AI practitioners advocate for a human in the loop to ensure important decisions aren't simply rubber-stamped, Workday's Nicole Hemler says the company's solution takes a more rigorous approach.

"We don't let agents take actions on behalf of people in talent without explicit human involvement. We're building it specifically to give people at least two options they have to look at. This forces them to engage cognitively."

Engaging people actively in the process also ensures that domain knowledge, which has been a casualty in organisations that have jumped to downsizing in the age of AI, is not lost.

The impact of incorrect decisions in HR can be significant. As well as engaging humans to make decisions at critical stages of different processes, the design of the processes automated through Workday's platform uses deterministic AI. This achieves two major benefits. Deterministic models learn what works and retain that in memory and reuse the same actions. This reduces operating costs as the AI doesn't need to recalculate the next steps it needs to take or suggest.

Where there are choices, probabilistic models are used with human intervention to ensure that the decisions undertaken are correct.

"The core of our AI thesis is to combine reasoning models with deterministic guardrails. That's really the heart of work today," explained Joel Hellerman, Workday's Chief AI Officer. "The very heart of our approach to not build rogue agents that even have a two percent probability of going off the rails."

Amongst the many challenges that come with implementing AI in HR is bias. Many AI models have been trained on data that has been shown to have cultural, gender and other biases. In recruitment and personnel management bias is not only unfair. It leaves organisations open to compliance issues that can lead to legal consequences. With many AI models lacking transparency, understanding whether a model has intrinsic biases, either through the design of algorithms or because of training data, is paramount.

With the acquisition of Paradox in October 2025, Adam Godson joined Workday and is now the General Manager of Talent Acquisition Products. He said Workday's approach is to be transparent.

"We publish bias testing results for our agents. We will continue to have openness as we think about all our agents in hiring to be sure that they don't have bias in those processes," he explained.

The steps Workday is taking come from looking at AI as being either lawful or lawless. The strong focus on governance, transparency and ensuring that humans in the loop do more than rubber stamp AI outputs ensure that governance is not only paid lip service. It makes governance an active part of the AI equation.

Perhaps the most valuable deliverable from the use of AI in HR is that it enables practitioners to elevate their work. AI can carry out repetitive tasks such as scheduling interviews, doing a first pass through job applications and, importantly, identifying talent and skills that already exist within organisations. Those can be painstaking tasks that take time away from HR operatives who could be spending time on high value tasks such as interviewing candidates, supporting education and creating professional development programs.

AI becomes an enabler that allows HR teams to spend more time on the human part of HR and less time on administrative tasks. However, this demands a strategic rethink in how HR teams are structured.

"There's a tremendous reskilling, upskilling and change effort that will be required in HR organisations," said Helmer. "What I hear from HR leaders is the need to change their own functions and processes. They don't want to do is automate what they are already doing today.  What they really want to do is rethink from the ground up and ask what are the skills they need and whether they have those skills."

From that point, it becomes a question of restructuring and educating teams to use AI to be more effective and enable HR to spend more time with people and less on repetitive and laborious processes.

For HR leaders, the priority is to adopt AI with clear governance, meaningful human oversight and a practical plan for workforce capability. Used this way, AI can improve efficiency while allowing HR teams to focus more on people, judgement and organisational needs.

Leveraging deterministic AI and rigorous bias testing, coupled with explicit human involvement in critical decisions, mitigates risk and ensures responsible application. This evolution represents a fundamental shift, transforming AI from a simple automation tool into a core enabler for HR's most impactful work.