The zero touch future: Enabling Telstra’s path to a fully autonomous, self-healing network

At MWC 2026, Telstra announced a major step forward in its journey towards building one of the world’s most advanced autonomous networks in collaboration with Red Hat and other industry partners. Telstra executed a proof-of-concept within a live telecommunication (telco) cloud environment to autonomously detect and resolve an unplanned infrastructure outage by shifting critical network applications to healthy infrastructure in just minutes. The result shows how a multivendor, AI-native architecture can enable self-healing operations at scale. Central to this exciting new capability are the open, scalable, and production‑grade AI and automation platforms from Red Hat.

Before explaining the new capability, let’s explore how telco networks currently manage disruptions, such as a failure in a key component of the technology stack.

In this scenario, an operations engineer often spends hours performing manual triage to answer critical questions:

• Is this a hardware or software issue?
• Is it related to the compute layer, storage layer, networking layer, application layer, or something else?
• If it's a software issue, which component is malfunctioning, and which specific commands are needed to ascertain the cause?
• What do I need to do to remediate this issue?

As demonstrated by Telstra, AI-assisted operations aim to significantly simplify remediation by taking advantage of both AI and automation—reducing recovery times from hours to minutes while increasing platform stability and elevating the customer experience.

The first phase of demonstration showed how gen AI can assist human operators by unifying multivendor data into an AI-enabled interface, in this case Red Hat OpenShift Lightspeed. The second demonstration showcased a fully autonomous, AI- and automation-enabled self-healing loop, allowing the AI agent to apply reasoning to select the best remediation approach.

Powered by Red Hat OpenShift AI, this agent enables cross-platform interactions, instantly communicating with third-party management platforms and Telstra’s Knowledge DB via MCP servers.

Once the AI agent determines the correct course of action, it automatically triggers Red Hat Ansible Automation Platform to apply the corrective action directly to the application. With Policy as Code (PaC), role-based access control (RBAC), and full audit trails, Ansible Automation Platform makes sure that every corrective action is authorized, consistent, and fully documented.

By using Red Hat OpenShift as the application platform, OpenShift AI as the intelligence layer, and Ansible Automation Platform as the trusted execution engine, Telstra is charting a path to autonomous self-healing networks. While the technological leap for network operators is profound, the true measure of Telstra’s AI-enabled autonomous network lies in the experience of the end customer. 

How AI self-healing protects the customer experience

Imagine a small business owner who is in the middle of a live-streamed product launch to thousands of online customers. She is relying on a rock-solid, high-speed 5G connection to broadcast her video feed. Deep within a Telstra datacenter miles away, a sudden hardware fault occurs. A critical server hosting the cloud-native network functions (CNFs) that process her mobile traffic begins to fail. In the past, this hardware failure would have immediate and visible consequences. She would lose her audience, lose momentum, and likely blame the network for the disruption. Meanwhile, an operations engineer would be scrambling to restore her service.

With the AI self-healing architecture in production, the outcome is different. The moment the hardware begins to degrade, the AI agent, powered by OpenShift AI, detects the anomaly and autonomously gets to work. It communicates with the underlying infrastructure, understands the context, and uses Ansible Automation Platform to shift network traffic to healthy systems.

The impact to the end customer is also very different. 

For the customer, the live stream disruption is only a minor inconvenience, and the audience continues to remain engaged. She finishes her product launch successfully, completely unaware that a critical piece of infrastructure just failed and was autonomously replaced in real time.

By removing the human from incident response, the AI-driven system detects and resolves network degradation before it significantly impacts the end customer. Telstra’s showcase at MWC is more than just an impressive technical demonstration; the PoC proves that the vision of a fully autonomous network is no longer a distant, theoretical concept—it’s both practical and achievable.