How managers and leaders can navigate AI’s workplace risks while pursuing its productivity potential
AI smart glasses could soon be restricted in Commonwealth workplaces, as the federal government responds to privacy and security concerns about devices capable of recording people without their knowledge.
The review comes as the government develops Australia’s first best-practice guidance for workplace AI. In September, Employment Minister Amanda Rishworth brought together Commonwealth Bank, Microsoft, Telstra, Kmart and other major employers to shape it.
The roundtable was one of five planned consultations with unions, employers, small businesses, employer peak bodies and academics examining how workplaces can realise AI’s productivity benefits while supporting workers through change.
“To realise the potential productivity gains from AI, workers and businesses need the capability to adopt new technologies with confidence that appropriate safeguards are in place,” Rishworth said.
While AI smart glasses are making headlines now, the debate over their use is just a precursor to the growing number of leadership tests organisations will face as the technology is pushed deeper into working life.
The Productivity Commission estimates AI could deliver $116 billion in benefits to Australia over the next decade – around $4,300 per person. But that opportunity comes with risks involving privacy, safety, data, accountability, trust and its effects on employees.
AI is one part of a broader reform agenda aimed at reversing Australia’s long-running productivity slowdown, with the latest ABS figures showing GDP per hour worked fell 0.2 per cent over the year to June 2026.
Who actually delivers Australia’s productivity agenda?
The role managers and leaders play in delivering Australia’s broader productivity reform agenda on the ground is the focus of the Institute of Managers and Leaders’ (IML) latest report, The delivery layer: Why Australia’s productivity agenda depends on organisational leadership.
IML mapped all 47 recommendations spanning the Productivity Commission’s five productivity inquiries and found that 32 rely significantly on organisational leadership for their delivery in workplaces across Australia.
The report then draws on insights from IML CEO Sam Bell, NSW Productivity and Equality Commissioner Peter Achterstraat AM and independent economist Saul Eslake to explore how that leadership dependence plays out in practice.
One recommendation brought that reliance into particularly sharp focus: the Commission’s position that AI-specific regulation should be a last resort.
Our mapping assessed it as requiring all four leadership capabilities drawn from the IML Competency Framework – driving organisational change, developing people, collaborating across boundaries and workforce planning.
That finding reflects a practical reality. AI is evolving too quickly and taking too many different forms for governments to prescribe how every application should be used in every workplace.
That puts the onus on managers and leaders to interpret a regulatory environment that is still taking shape and establish appropriate safeguards inside their organisations.
As Bell argues in the report, AI adoption needs to be deliberate rather than left to emerge unevenly. Without a clear internal framework, organisations risk a “randomised” application that invites poor practice.
That uncertainty does not mean organisations are operating in a legal vacuum. As Achterstraat highlights, privacy and data laws already apply to AI use, and leaders who fail to govern its application risk real harm – not just to their organisations, but to people affected by automated decisions.
Looking at productivity more broadly, Eslake argues that government can establish incentives and frameworks, but outcomes ultimately depend on the decisions of those operating within them.
What does successful AI adoption demand from leaders?
For managers and leaders, realising AI’s productivity potential increasingly calls for adaptive leadership – an approach that recognises AI adoption as an evolving organisational challenge rather than a one-off technology decision.
IML’s April report, Adaptive Leadership in the AI Era, found that meeting this challenge depends less on technical proficiency than on judgement, contextual thinking and the ability to lead people through uncertainty and change.
Unlike a technical problem with a known solution, an adaptive challenge requires people to work out new ways of operating as conditions shift. IML’s Observe, Interpret, Respond and Reflect framework gives managers a practical way to do that: understand what is happening in the workplace, interpret the context, act accordingly, then learn and adjust.
That framework can be seen in practice in Benetas CEO Sandra Hills’ account of introducing Abi, a companion robot developed by Australian robotics firm Andromeda, across two aged-care homes. It was new territory for everyone, which left some employees unsettled about what it meant for their roles. When the program expanded to a second home, Benetas applied those learnings by engaging key stakeholders before Abi arrived and involving staff in shaping and evaluating its integration. The robot was the same – the difference was adaptive leadership in action.
What happens as AI – and the rules around it – keep changing?
As AI continues to evolve, governments will keep adjusting laws, policy and guidance in response to its opportunities and risks. Those changes will flow through to organisations, where managers and leaders will need to adjust workplace policies and ways of working, guide AI adoption and manage its risks to employees and the organisation – all while seeking to maximise its productivity potential.
Explore the exact leadership capabilities successful AI adoption will demand – and why managers and leaders are critical to delivering Australia’s wider productivity agenda.
