•
Jul 22
Sand Technologies
Telecom operators have long used automation to manage scale, from spotting network faults to handling customer service issues. However, today’s networks are more distributed, services are more complex and customers expect faster, more personalized services. Traditional automation, built on static rules and fixed workflows, is no longer enough to keep pace with the fast-moving, complex world of modern telecom.
The rise of Agentic AI in Telecom is enabling providers to respond more effectively to today’s operational and customer demands. These AI systems go beyond automating tasks. They’re designed to work towards goals, make decisions and take action within clear boundaries. For telecom, that means a future of more autonomous networks, smarter operations and new ways to deliver value to customers.
Telecom sits at the heart of today’s digital economy, powering connectivity, enhancing productivity and driving innovation across industries. It’s a rapidly growing market, expected to reach $2.87 trillion by 2030. But despite this growth, the sector faces increasing operational and strategic challenges.
Networks are becoming more distributed across different technologies. From 5G and private networks to edge computing, connectivity now reaches deeper into businesses, cities and devices, creating more complex and distributed ecosystems. Telecom services have also expanded to include enterprise IoT, digital platforms and API monetization. This expansion introduces new layers of complexity to the delivery and management of services.
At the same time, customer expectations have evolved. People now expect fast, personalized and always-on services as the standard. Meeting these expectations is more critical than ever, as competition grows and customer acquisition becomes more expensive.
To manage this complexity, telecom providers have long relied on automation. But traditional automation follows fixed rules; it can optimize processes, but it can’t adapt to change or anticipate new needs. Ultimately, it has become clear that for telcos to keep up, they need a new approach. One that allows systems to operate with autonomy and continuously adapt to evolving conditions – something that agentic AI can offer.
In telecom, Agentic AI isn’t a single solution. It’s a capability layer that spans domains, embedding intelligence where action is needed most. From network performance to customer service, agents act with purpose, diagnosing, deciding and executing within business-defined constraints. Here are a few ways it’s being applied:
| Domain | Agentic AI in Action | Impact |
| Network Operations | AI agents dynamically reroute traffic, preemptively address faults and optimize energy use in real-time. | Higher resilience, reduced OPEX |
| Customer Experience | Agents resolve complex queries across billing, service and technical systems without requiring human handoffs. | Faster resolution, improved NPS |
| Service Monetization | Agents dynamically adjust offers, pricing or network resources based on behavior and business goals. | Increased ARPU, reduced churn |
| Operational Efficiency | Agents orchestrate cross-department workflows, bridging silos between field, service and network operations. | Greater agility, reduced delays |
The above examples reflect a broader shift that’s taking place in the telecom industry. At its core, this shift is about driving greater efficiency, not just to reduce costs, but to enable the speed, flexibility and intelligence needed to stay competitive in a rapidly evolving market.
But the case for Agentic AI goes beyond efficiency alone. It’s also about building resilience, creating differentiation and fueling growth through the following ways:
Agentic AI allows networks to shift from simply reacting to problems toward actively preventing them. It monitors conditions in real-time and makes automatic adjustments, such as rerouting traffic or fixing issues, without waiting for human intervention. This action helps maintain stability, improves resilience and keeps services running smoothly.
Telkomsel, one of Indonesia’s leading digital telecommunications providers, is already applying agentic AI in this way. Its award-winning solution forecasts future network demand and suggests site-level adjustments before congestion or service issues arise. By anticipating problems early, the technology allows operators to keep networks performing optimally without waiting for issues to appear.
By handling complexity behind the scenes, AI agents free human teams to focus on higher-value work. This enables staff to devote more time to addressing complex issues, providing personalized service and engaging in meaningful customer interactions. The result is faster resolutions, smarter personalization and stronger customer relationships.
Vodafone Business, in partnership with ServiceNow, offers a clear example of this in action. Its agentic AI platform proactively detects and resolves service issues within minutes, while integrated AI and digital channels ensure customers are routed to the right team first time. This approach frees human staff to focus on complex, high-value interactions. It has led to a 45% increase in digital engagement and a 4x improvement in customer satisfaction.
Agentic systems open the door to more adaptive and flexible services that help operators move beyond simply selling fixed capacity. With these services, telcos can offer on-demand services, such as bandwidth that scales in real-time based on usage, or enable dynamic API marketplaces that allow partners and customers to tap into network capabilities as needed. This creates new, more responsive business models where agility itself becomes a revenue stream.
Google Cloud’s new agentic AI platform brings this to life. Through its Agent-to-Agent protocol and AI Agent Marketplace, telecom operators can deploy specialized AI agents that communicate, coordinate and act autonomously. These agents can detect network opportunities or issues in real-time and automatically trigger new services, such as activating additional bandwidth or enabling edge computing when demand spikes. This turns network agility into a revenue stream, enabling operators to offer more responsive, on-demand services beyond traditional fixed capacity.
Realizing Agentic AI’s full potential in telecom is not just about deploying LLMs or chatbots. It requires a combination of specific, strategic technologies working together:
Together, these components provide the framework that allows telecom systems to advance safely and autonomously.
Bringing autonomy to critical infrastructure, such as telecom, comes with real responsibility. As AI takes on more decision-making responsibilities, telecom leaders need to establish clear boundaries and safeguards to ensure these systems operate in the right way.
This starts with defining decision limits, such as financial thresholds or actions that could impact service levels, so AI agents know where human approval is still required. Regulatory acts such as the EU AI Act will be key in navigating this. Starting in August 2026, operators of critical infrastructure and high-compute AI systems will likely be required to ensure human oversight and retain the ability to override AI decisions.
Transparency is also key. AI decisions should be both traceable and explainable, allowing teams to understand how outcomes were reached and maintain trust with both customers and regulators.
Finally, oversight needs to evolve. Moving toward a ‘human-in-the-loop’ model allows people to supervise AI without slowing it down with constant approvals. This creates a balance where AI drives efficiency, but human judgment remains in place where it matters most.
For telecom, the real opportunity with Agentic AI isn’t about chasing trends. It’s about solving practical, persistent challenges: how to run networks more efficiently, how to meet rising customer expectations without rising costs, and how to create services that are flexible enough to evolve alongside the market.
The next move for telecom providers is to shift from pilots to meaningful integration. Embedding Agentic AI into core systems will enable faster, more informed decision-making, reduce inefficiencies, and open up new opportunities for service innovation. Ultimately, it’s not about adopting AI for its own sake. It’s about building smarter, more adaptive operations that keep pace with the demands of a changing industry.
Other articles that may interest you
