EQUIFAX ANZ
Rather than getting caught in hype cycles, Paul focuses on practical applications with clear objectives.
“ I break the conversation down into three things because it impacts humans,” he explains.“ Are you assisting a human? Are you augmenting a human? Or are you transforming an activity?”
This framework helps Equifax determine appropriate risk levels and implementation strategies.
For instance, all developers at Equifax have access to Microsoft Copilot to boost productivity, an approach Paul classifies as“ assist”.
“ When I first rolled it out, the engineers said to me,‘ Does that mean we’ re going to downsize because you can now do more?’ I said,‘ No, now I can get more of a backlog done’.”
This productivity boost has been significant, with Paul noting:“ My technology team shouldn’ t be 650 people, it should be about 1,200 or 1,300 people based upon the workload we actually get done. I put 1,700 changes into production a year plus.”
The company has also embraced a crowdsourced approach to exploring more advanced AI applications.“ I had 70 people volunteer to work on agentic AI,” Paul says.“ They said,‘ we want to do it, we want to work on it, we want to manage it.’”
This structured yet grassroots approach to AI innovation balances enthusiasm with prudence.
“ Agentic AI in its first instance will be an augment only,” Paul emphasises.
“ A human has to be in the loop. At some point in time in our future, we may well get mature enough that we want to call it a transform and the human is removed. But on day one, we’ re only at the augment stage.”
This grassroots enthusiasm reflects a culture where technological innovation is embraced and understood throughout the organisation.
As Paul notes:“ The workforce is really engaged because they read about agentic AI in the paper and their mates talk about,‘ I wonder if it’ s ever going to come to my organisation,’ they sit there saying,‘ What do you mean? We’ re already doing it.’”