Ericsson is reshaping the OSS/BSS modernization debate around business results. The company now argues that AI must support measurable operator gains, not just technical upgrades.
This shift comes as communications service providers pursue more autonomous networks. These networks can make more decisions with less human intervention. However, they need reliable data, trusted systems, and clear operating goals.
For Robert Curran, consulting analyst with Appledore Research, the market faces a defining question. “The first and most critical one, I think, is just the impact of AI in this whole area and understanding…where it should have an impact, where it can have an impact, what makes most sense,” he said.
At the same time, operators must decide which use cases deserve early investment. Some can improve customer care, service launch, or network operations quickly. Others may remain interesting, but not yet urgent.
Data remains the harder issue. “Data is very fragmented,” Curran said. “It’s still the case — I know it’s a cliché, but it’s a cliché because it’s true— which is the siloed data.”
That fragmentation can slow automation projects. Autonomous systems need timely information from different business and network domains. If data stays trapped in separate systems, AI decisions become less reliable.
Jason Keane, Ericsson Head of Portfolio Business and Operations Support Systems, sees the same challenge. “AI is a huge impact on the industry, but if you don’t have good data, you’re really not getting a good outcome,” he said.
He also warned against AI spending without visible returns. “We need to get down and get the actual value back out of the network.”
Ericsson’s answer is its Business Value Pathways approach. It starts with specific business priorities, then maps the required technology. The focus areas include monetization, customer experience, and operational efficiency.
The announced pathways cover faster data transformation and zero-touch product launch. They also include agentic AI service experience and intelligent IT operations. Agentic AI means software agents can act toward goals, under defined controls.
This model may help operators avoid isolated automation projects. It can also link network modernization to revenue growth and customer satisfaction. Yet success still depends on governance, clean data, and organizational trust.
Keane said Ericsson wants to focus on “the actual impactful use cases that we can deliver.” He added, “We probably spend a little too much focus on the wonderful technology but not the business outcome we want.”
Agentic AI also changes how suppliers design OSS/BSS platforms. It should not sit outside daily operations. “It’s not just that you bolt AI on the side, but you embed it in what you do,” Keane said.
Curran sees this as a wider market movement. “AI is everywhere, agentic is…everywhere in the OSS/BSS portfolio,” he said.
For CSPs, the message is direct. Autonomy will not arrive through hype alone. It will require trusted intelligence, connected data, and stronger links between technology and business value.
