Agentic AI accelerating the journey to autonomous networks

What if telecom networks could think, analyse, and act without human intervention? Autonomous networks are not just a dream—they are quickly becoming necessary in today's hyper-connected digital world. As telecom continues to evolve, one technology stands out as a game-changer in realising this vision: Agentic AI. Agentic AI is poised to accelerate the industry's transformation into fully autonomous operations by empowering networks to make complex, real-time decisions across multivendor ecosystems. 

The Promise of Autonomous Networks in Telecom 

Autonomous networks represent the pinnacle of telecom automation. They promise intelligent, real-time decision-making across the entire network lifecycle—from planning to operations. At its highest level of autonomy, Level 5 (L5), a telco network, transforms into a self-driving entity requiring minimal human intervention. A recent CSIRO report indicates that 68% of Australian businesses have implemented AI technologies, with an additional 23% planning to do so within the next 12 months. 

This level of automation redefines the efficiency of network operations, reduces manual errors, slashes operational costs (OPEX), and significantly improves customer experience through proactive service management. 

The autonomous networks market is growing rapidly, projected to reach $17.5 billion by 2029, with a  CAGR of 20.1%. However, despite this growth, 80% of telcos remain stuck at Level 2 autonomy. Legacy systems, siloed technology domains, and poor-quality data continue to block their path forward. 

The breakthrough lies in adopting specialised technologies, including Agentic AI, to overcome these challenges one domain at a time. 

Agentic AI—The Brain Behind Smarter Networks 

Agentic AI serves as the orchestrator that seamlessly integrates the diverse components of an autonomous network. It connects specialised AI models and algorithms from network domains to enable cohesive, real-time decision-making. Using Agentic AI, networks can transition from reactive to proactive and predictive, unlocking capabilities such as self-healing, self-optimisation, and agile service provisioning. 

Agentic AI is expected to revolutionise the Australian workplace by 2025, potentially addressing productivity challenges and mitigating employee burnout. With a market size poised to grow at a CAGR of 19.4%, from $10 billion in 2023 to $48.5 billion by 2032, Agentic AI's value proposition for telecom is becoming impossible to ignore. 

Breaking Down the Use Cases for Agentic AI 

According to a study, Adopting agentic AI could significantly boost Australia's economy, with projections indicating contributions of up to $45 billion annually by 2030, depending on the pace of adoption. To fully harness the potential of Agentic AI, it must be deployed thoughtfully across the lifecycle of a network.  Here's how it can be applied in key use cases: 

1. On-Demand Network Planning & Design 

Traditional network planning involves manual or semi-automated processes that cannot meet demand. Agentic AI enables predictive, real-time planning by connecting multiple specialised agent functions, such as virtual resource planning and infrastructure design. These agents analyse data from each domain and work collaboratively to deliver end-to-end network plans tailored to operational needs.

2. Real-Time Network Build 

While physical infrastructure remains integral to network builds, autonomous networks demand dynamic, real-time adjustments to virtual nodes and configurations. Agentic AI leverages generative AI  to determine optimal paths, perform intelligent routing, and generate real-time configuration data,  reducing deployment times and ensuring seamless service activation. 

3. Digital Twin for Network Change Management 

Change management is one of the most challenging aspects of telecom operations. Agentic AI  facilitates the creation of real-time digital twins, replicating the network's current state to assess the impact of proposed changes. This eliminates guesswork and ensures changes are implemented with minimal disruption to ongoing services. 

4. Dynamic Network Service Provisioning 

Agentic AI automates the traditionally static service provisioning process. Service creation,  configuration, and activation occur dynamically in real time, enabling telcos to bundle and deploy services faster—often within minutes rather than days. 

5. Closed-loop assurance and Zero-Touch Operations 

The holy grail of autonomous networks lies in their ability to self-heal and self-optimise. Agentic AI  integrates fault correlation, impact analysis, root cause identification, and action planning to create closed-loop systems—networks that can fix issues without human intervention. 

Tackling Industry Hurdles with Agentic AI 

The road to full autonomy isn't without its challenges. Here's how Agentic AI addresses some of the key  hurdles faced by telcos:
 

• Integration Into Legacy Systems: Agentic AI enables interoperability across multivendor ecosystems by harmonising and orchestrating intelligence from diverse systems. It ensures that telcos can upgrade incrementally, avoiding costly rip-and-replace strategies. 

• Workforce Adaptation and Resistance: Transitioning to autonomous networks requires upskilling existing employees. Agentic AI's granular automation allows telcos to balance human oversight and machine-led operations while gradually phasing in autonomy. 

• Data Quality and Security: Agentic AI's federated architecture supports real-time data analysis while adhering to stringent data governance standards. It ensures data quality and security even within complex, multi-stakeholder ecosystems. 

• Regulatory and Compliance Challenges: The telecom industry needs updated frameworks to accommodate autonomous systems. Collaborative efforts like the TM Forum's work on standardising intent-driven interfaces are paving the way. Agentic AI plays a critical role in realising these standards. 

A Path to Ecosystem Collaboration 

The future of telecom lies in deeper ecosystem collaboration. With networks increasingly dependent on multivendor components and AI models, the Linux Foundation's networking initiatives and open collaborations through TM Forum are paving the way for unified frameworks.

Agentic AI acts as a bridge between proprietary vendor-specific AI models, enabling cross-domain analysis and facilitating ecosystem-wide collaboration. These capabilities ensure that telcos can achieve end-to-end connectivity and deliver on the promise of fully autonomous networks. 

Why Agentic AI Is the Answer 

Autonomous networks are rapidly reshaping telecom, and Agentic AI accelerates this transformation. According to research, in Australia, adopting agentic AI is anticipated to revolutionise workplaces by 2025, potentially addressing challenges such as low productivity growth and employee burnout. 

By stitching specialised AI functions together, enabling bright orchestration, and providing actionable insights, Agentic AI is closing the gap to full autonomy and creating a new paradigm for the telecom industry. 

The benefits are clear. Telcos adopting Agentic AI can achieve the following: 

• Improved SLA adherence 

• 20–30% OPEX savings 

• Rapid service deployment and enhanced customer experience