MegaFon estimates: up to 40% of telecom operators' routine tasks will shift to AI within three years
MegaFon predicts that within three years AI agents will be able to perform up to 40% of standard service and operational tasks at telecom operator companies. The company clarifies: this isn't about all functions, but about repeatable operations where AI agents and classic automation tools can be applied simultaneously. Managing Director for Digital Solutions Evgeny Bobkov emphasized that autonomy does not mean a complete abandonment of human control. Systems are already handling specific tasks today: • Collecting diagnostic data • Merging related incident events • Creating and routing tickets • Executing pre-approved scenarios with result verification and rollback capability
AI-processed from CNews; edited by Hamidun News
MegaFon's Managing Director for Digital Solutions, Evgeny Bobkov, offered a forecast for the telecommunications industry during a session at the "Digital Solutions" forum. Within three years, up to 40% of standard service and operational tasks at telecom operators could be performed without human intervention — provided these are repeatable operations where AI agents and classic automation tools can be applied simultaneously.
"40% is a scenario-based estimate of the share of standard service and operational tasks performed end-to-end without human intervention," Bobkov clarified. The company is not talking about automating all functions entirely, but specifically about repeatable operations where AI and traditional automation complement each other.
What can AI agents already do today?
Bobkov listed specific tasks that can already be handed off to agents today:
• Collecting diagnostic data — automatically gathering information about the state of systems and networks • Merging related incident events — analyzing incidents and identifying their correlation • Creating and routing tickets — generating tickets and distributing them to the necessary specialists • Executing pre-approved scenarios — carrying out previously approved sequences of actions with result verification and rollback capability
Human control remains a priority
A critical clarification: the autonomy of AI agents does not mean a complete removal of humans from the system. Control over the system must remain in human hands. "It's possible to have no human in every operation, but it's not possible to have no human control over the system," Bobkov said.
The company also does not recommend shifting everything to AI indiscriminately. If a task can be fully described by rules and algorithms, it is often cheaper and more reliable to use classic automation tools than to train a neural network. Thus, MegaFon sees the future as hybrid, where AI agents and traditional automation work within the same system, each in its proper place.
MegaFon's forecast reflects a broader trend in the telecommunications industry: operators are actively experimenting with AI to optimize operational processes, while preserving the critical human role in managing and overseeing the system. This allows them not only to boost efficiency but also to avoid losing control over complex engineering processes.
Key Questions Answered
What can AI agents already do today?
Bobkov listed specific tasks that can already be handed off to agents today: Collecting diagnostic data — automatically gathering information about the state of systems and networks Merging related incident events — analyzing incidents and identifying their correlation Creating and routing tickets — generating tickets and distributing them to the necessary specialists Executing pre-approved scenarios — carrying out previously approved sequences of actions with result verification and rollback capability
What are the key details on Human control remains a priority?
A critical clarification: the autonomy of AI agents does not mean a complete removal of humans from the system. Control over the system must remain in human hands. "It's possible to have no human in every operation, but it's not possible to have no human control over the system," Bobkov said. The company also does not recommend shifting everything to AI indiscriminately. If a task can be fully described by rules and algorithms, it is often cheaper and more reliable to use classic automation tools than to train a neural network. Thus, MegaFon sees the future as hybrid, where
Delegate Deep Research, Coding & Multi-Step Workflows to Nomad Agent
Deploy autonomous task execution with Nomad Agent (agent.nomadnet.ai) — automated web research, data synthesis, code generation, and MCP tool orchestration.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.