HR Leader Podcast Network

The episode covers legitimate AI risk topics (automation bias, AI as decision-maker vs. assistant, human-in-the-loop, investigation misuse) with some depth, but relies heavily on restating the same core principle throughout. The discussion of specific workplace risks is useful but repetitive; most insights cluster around one thesis. Several soft-pitched questions allow the guest to reiterate frameworks without being pushed deeper into novel territory or edge cases.

“AI as a decision maker is one of the topics that is going to have increased attention going forward”

“where we need to be cautious is this shift in, or movement between using it as an administrative assistant or an assistant to get through large volumes of documents or materials and understanding quickly the state of play to outsourcing entirely your judgment or delegating tasks to AI to complete without any human oversight”

The frameworks presented - human-in-the-loop, automation bias, AI hallucination, de-identification - are well-established in AI governance discourse and circulate widely in compliance and legal circles. The application to Fair Work Act and Australian employment law adds minor geographic specificity, but the core insights are not contrarian or first-principles. No counterintuitive positions or novel risk framings are offered.

“if you would not be comfortable providing this piece of information to a stranger on the street, don't put into AI”

“AI can hallucinate, AI can get things wrong and it can misinterpret information”

Michael Stutley is a workplace law partner with relevant domain expertise and appears to advise on employment matters daily. However, the transcript provides no evidence of specific scale (clients managed, landmark cases, direct AI incident response experience). He speaks as a practitioner with real exposure to workplace investigations and Fair Work issues, but lacks credentials demonstrating he has dealt with high-stakes AI deployment failures or worked across multiple sectors. Solid but not exceptional caliber for this topic.

“I'm a partner at Kingston Reed. I'm based in the Perth office. I spend quite a bit of time in our Brisbane office as well.”

“largely the practice is centered around employment, uh, workplace relations, industrial relations, safety and global mobility”

The episode is notably light on named examples, quantified risks, or real case citations. Medical reports and investigations are mentioned as risk areas, but no specific Fair Work Commission cases, employer AI incidents, or data on investigation errors from AI are referenced. The golden rules are prescriptive but abstract (de-identify, preserve source docs, apply newspaper test) with minimal concrete scenarios showing failure modes. The lack of 'here's what happened when...' damages credibility.

“Medical reports are a really good example of that”

“if you inputted a person's medical report into... who has control of that information? Where does it go?”

Jerome asks competent setup questions and demonstrates active listening (e.g., 'so just to clarify then'), but rarely pushes back or probes beyond the guest's initial answer. When Michael makes broad claims (e.g., 'AI can hallucinate'), Jerome moves on rather than asking for examples. Questions are generally softball and affirming, allowing the guest to deliver prepared talking points without friction. There is one strong follow-up about worker pressure to use AI, but overall the conversation lacks the rigor needed to stress-test claims or explore nuance.

“Tell me about what you're seeing as some of the hidden or perhaps very visible risks, uh, from AI use across the workplace right now”

“there's a lot of societal or cultural pressure for workers to be using AI right now, whether that's intentional or not... it seems like it would be a really hard balance to strike”