Ericsson and Telstra are advancing a new model for autonomous telecom networks. Their collaboration places intent, knowledge, and trusted AI at the center of operations.
The companies first announced the partnership in 2025. Since then, they have focused on practical use cases. They also address deeper obstacles in modern networks.
These include fragmented data, mixed vendor environments, and limited trust in AI decisions. Operators also face a persistent gap between business goals and network execution.
Mats Karlsson, Ericsson’s Head of Solution Area Business & Operations Support Systems, described the core idea clearly. “The central proposition is really that intent and the concept of knowledge [are] two fundamental aspects to going towards autonomous network,” he said.
This marks a shift from traditional automation. Today, many systems still follow human-built workflows. Engineers define each step before software executes the task.
Autonomous networks aim to change that model. Instead of defining every step, operators state the desired result. AI agents then plan, validate, and execute suitable actions.
Karlsson explained the difference directly. “Automation is really still human-defined flows and human-defined configuration,” he said. “When you’re going towards autonomous network, you’re actually telling the network what do you want the network to achieve, what outcome, what is the intent.”
For operators, the possible efficiency gains look attractive. Faster service launches could improve competitiveness. Better network response could reduce manual intervention. Smarter operations may also improve customer experience.
However, reliability remains the key concern. Telecom networks carry essential business and consumer services. Operators cannot allow uncertain AI behavior to disrupt live operations.
Mark Sanders, Telstra’s chief architect, framed this challenge in simple terms. “This is our network. This is our core business,” he said. “The opportunity is incredible…We’ve got to be able to do it with confidence.”
That confidence depends on what both companies call the knowledge plane. This concept gives machines the context behind network data. It connects relationships, policies, intent, and real network conditions.
Sanders summarized its value. “Knowledge is the relationships,” he said. “But most importantly, we’ve coded those relationships…into the actual data…So now that the knowledge can be read by machines, they can reason over and they can act.”
In practice, this could help networks avoid blind automation. AI agents would not simply guess the next action. They would reason using structured knowledge and defined limits.
Ericsson also links this vision to its OSS/BSS strategy. The company wants operators to connect autonomy with monetization, service quality, and efficiency. Its business value pathways target data transformation and zero-touch launches.
Still, the transition will not happen overnight. Operators need targeted projects before broader rollout. Vendors and standards groups must also align on rules and responsibilities.
Karlsson underlined that point. “This will require a full industry collaboration,” he said.
For telecom engineers, the message is clear. Autonomous networks are not only about adding AI. They require trusted data, clear intent, and disciplined governance. Engineers will increasingly guide the system, rather than control every step.
