The Organisational Disciplines Behind Effective AI Adoption

Our Human-Led, AI-Enabled event brought together senior leaders from Coles, Qantas, and Accenture to share what responsible, practical AI adoption actually looks like inside complex organisations. The patterns across their very different contexts were clear enough to draw some direct conclusions. 

Start with the problem 

Every productive AI adoption story starts the same way, with a clearly defined problem, and an honest assessment of what it costs to leave it unsolved. 

Coles identified 20,000 to 30,000 hours of annual capacity tied up in self-serviceable HR queries. Qantas had 130,000 safety reports a year containing signal that no human team could reliably surface at scale. Accenture found a processing team spending 15% of their time on jobs that couldn't be completed due to missing data. In each case, the problem was defined before the technology was selected and the solution was proportionate to it. 

Davin d'Silva put it plainly: sometimes AI is the right tool, and sometimes a macro, a process redesign, or simply stopping a service altogether gets you further faster. The pressure to find an AI use case can lead organisations toward solutions that introduce risk and tech debt without matching value. Executives have a responsibility to filter that pressure for their teams. 

Governance is a leadership decision 

Mark Lipman shared Qantas's explicit position: generative AI will not be used to make or assist in making employment decisions, including recruitment, hiring, retention, promotion, performance monitoring, discipline, demotion, or termination. That kind of clarity is a leadership choice, and it matters. 

Kelly Brough noted that many organisations have responsible AI policies their leadership isn't actively close to. As AI becomes more embedded across operations, those policies need to be living documents, regularly reviewed, clearly owned, and specific enough to cover the real decisions being made day to day. 

The workforce shift is structural 

The more significant strategic question is what AI means for the capability profile of the workforce and how talent strategy needs to respond. 

Kelly framed this as reinventing not just the work, but the workforce that aligns to it. Davin was direct: leaders have historically been promoted and then loaded with administrative work. AI creates the conditions to change that, freeing leaders to focus on what actually distinguishes good leadership: decision-making, empathy, and developing their people. 

"The leaders of tomorrow need decision-making, empathy, and coaching. AI will push off the administrative load.

Our job is to hire and develop for what's left." 

Davin D’Silva 

Mark raised a related question worth sitting with: as early-career professionals do less of the administrative work that has historically built professional judgment, how is that judgment cultivated instead? It's a talent pipeline design question, and one that needs deliberate attention rather than assumption. 

Connect the pilots 

Kelly identified a dynamic common across many organisations right now: a portfolio of successful pilots that haven't yet added up to meaningful transformation. The value compounds when those pilots are connected across an end-to-end process, which requires executive sponsorship, cross-functional governance, and genuine change management capability. 

Most organisations are still working through this. The successful ones will need to invest as heavily in governance, workforce design, data foundations and continuous learning as the technology itself. 

The talent lens 

The organisations best positioned for the next decade are building the capability now to scale AI responsibly. At The Next Step and The Safe Step, it's exactly the kind of question we work through with clients: what does an AI-ready workforce actually look like, and how do you build toward it?