In the first article in this series, we explored the potential of agentic AI for telecom operators. These autonomous systems can handle complex, multi-step tasks dynamically, integrating data from diverse sources to optimize workflows across networks, customer service, and operational processes.
Although still in its early stages, agentic AI is already delivering measurable improvements—such as faster decision-making and enhanced service reliability—surpassing traditional automation and data-driven approaches. Of course, deployment comes with risks and limitations, including data privacy and regulatory compliance, which we’ll examine in a later article.
Forward-thinking telcos are beginning to deploy agentic AI in live environments. Now is the time for business leaders to identify applications that can deliver impact today. Current use cases—from autonomous customer service and fraud prevention—mirror existing human roles, aiming to augment efforts and streamline operations.
Anomaly detection
Telcos have long relied on real-time monitoring and predictive analytics to detect anomalies and prevent downtime. Agentic AI takes this further by enabling self-healing, self-optimizing, and self-configuring networks. AI agents ingest telemetry data, detect unexpected events such as traffic spikes, and autonomously implement fixes—like reallocating capacity. By continuously analyzing vast network traffic data, they identify patterns and irregularities that humans might miss, enabling instant responses and root-cause analysis.
AI-powered customer service
Customer experience has evolved with bots and generative AI, but agentic AI goes further. It doesn’t just provide information or suggest next steps—it takes action on behalf of customers or service agents. Gartner predicts that by 2029, agentic AI will autonomously resolve 80% of common customer service issues. Progress is already visible: AT&T reports a 50% reduction in unanswered queries, while Vodafone handles 45 million interactions monthly with a 70% resolution rate—without human intervention.
Real-time support for field technicians
AI agents assist technicians with complex repairs by integrating real-time data with technical manuals and procedures. They provide step-by-step troubleshooting and can implement updates directly. This optimizes field operations, allowing technicians to focus on the most complex issues. Some telcos report a 25% reduction in repeat site visits and a 29% faster mean time to repair (MTTR).
Fighting fraud
AI has long been central to fraud prevention, but attackers now use AI to increase the scale and complexity of threats. Agentic AI strengthens defenses by autonomously detecting, predicting, and mitigating zero-day vulnerabilities and advanced persistent threats. These agents continuously analyze attack behaviors and adapt security profiles in real time, reducing the burden on cybersecurity teams and enabling them to focus on strategic tasks.
Optimised energy and financial planning
AI and data centers are often criticized for high energy consumption, but agentic AI can help address this challenge. Radio access networks (RANs) account for 75–80% of operators’ grid power usage, making efficiency critical. Agentic AI uses pattern recognition and predictive analytics to optimize energy consumption, dynamically adjusting RANs based on real-time traffic demand.
Beyond technical operations, agentic AI supports CFOs with financial oversight and forecasting. It enables scenario-based planning to simulate product costs and revenue projections. Microsoft reports a 15–20% improvement in forecast accuracy using its CoPilot tools.
Now is the time to build the foundations for agentic AI
Agentic AI is already proving its value across operations, customer service, and security, delivering measurable ROI. With AI investment accelerating in telecom, delaying deployment is no longer an option.
Telco leaders must act quickly, but speed alone isn’t enough. Success depends on building the right foundations—especially around data. For AI agents to perform effectively, they require data that is accurate, consistent, and timely. The structure and integrity of this data are critical—without it, deployments cannot succeed. Building semantically rich, reliable data foundations is the next essential step to unlock agentic AI’s full potential.
In the next article, we’ll outline these foundational requirements and share how Teradata is partnering with communications service providers (CSPs) worldwide to build the data infrastructures that enable trusted agentic AI adoption. As the engine behind many analytics functions, Teradata is uniquely positioned to help telcos transition to the agentic economy.