NBN Co AI could give RSPs more network intelligence

NBN Co’s AI strategy is moving toward an agentic phase that could see the network wholesaler open its data to retailers and give them greater visibility into its network.

The national network builder’s operations general manager Richard Halliday told attendees at the AWS Summit in Sydney that, to date, NBN Co has taken a slower approach with AI technology, being wary of exposing large amount of data to it.

However, he said NBN Co is now ready to start exploring agentic applications - and the ability to make more of its network data available to retailers is emerging as promising business case for the technology.

“We're at the point now where we're trying to move that [AI strategy] into agentic, looking at ways that we can be predictive, the way that we can see into the network and fault-find and do that automatically and effectively,” Halliday said.

As a result of that, NBN Co may soon be sharing more information about its network with retailers, he said.

“We don't do that without being cautious with regard to the way we approach that ... but where we see this emerging is that, we are a [network] wholesaler through our retail providers, so how we can use the capability to expose more of the network data to retailers and the broader industry to bring the whole industry along on that journey with us,” he said.

Halliday also revealed that, as one the country’s biggest providers of critical infrastructure, NBN Co is exploring way to use the technology to make its network more resilient.

“Our aim is to make our network more predictive, self-healing and adaptive, and we see the use of AI in those environments making that possible. That then serves our customers and then through our retail providers to the end customers across the nation,” Halliday said.

NBN Co still needs to be cautious with its approach to AI given the regulatory setting within which the network builder operates.

That, according to Halliday, means that the company has adopted a “dual-speed” approach to AI trying to strike a balance between moving quickly on innovation while minimising risk to national, critical infrastructure.

“We need to be very conscious with regard to how we expose anything that we do across the broad industry. We also need to be very conscious of [taking] a dual speed type approach to how we go about developing AI, where our core capabilities need to be maintained and maintained with high resilience and capability, whereas you can do far more innovation at the edge," he said.

“We are looking at, how do we actually manage and balance between those two functions."