Issue Brief: AI-driven telecom networks

Telecom network economics are under structural long-term pressure. Network complexity continues to rise, customer expectations for performance remain high, and cost pressures arise in the context of an expected data traffic growth deceleration in developed markets. This article is a collaborative effort by Gustav Grundin, Sebastian Cubela, and Tomás Lajous, with Borja Belda, Matyas Zetek, and Sebastián González, representing views from McKinsey’s Technology, Media & Telecommunications Practice. The rise of AI provides multiple opportunities for telcos to turn the tide. For starters, AI can be embedded in telco value propositions offering differentiated services. Operators can also participate in the emerging AI value chain by becoming the AI infrastructure backbone. And telcos can leverage AI to transform their operations from the ground up. Most directly, in an industry where improved operational efficiency has become central to the networks’ value equation, AI represents a rare opportunity to reimagine the domain and reset its economics, while delivering exceptional performance. According to our latest survey of global telco top executives, the network domain will be one of the primary focus areas for AI deployments during the next two years, alongside customer care. However, the difference between marginal gains and structural impact lies not in the technology itself but in how operators redesign processes, roles, governance, and budgeting around it. To capture these opportunities, operators need to raise their ambition and apply the same type of discipline to the network domain. Recent progress in this field is encouraging. One of the loftiest and potentially most consequential ambitions many operators have articulated—a fully autonomous, self-optimizing, self-healing network—is no longer a distant vision, but an achievable goal in the coming years. AI’s impact on networks spans both capital expenditure (capex) and operating expenditure (opex), across the following three major domains. Network planning is shifting from static engineering thresholds to AI-driven, value-based optimization. To complement traditional network planning based on capacity and coverage estimates, advanced machine learning models and digital twins now simulate thousands of rollout and upgrade scenarios before capital is deployed. These simulations consider and estimate impacts on customer experience (CX), traffic evolution, customer churn and average revenue per user (ARPU), and potential competitive moves from other operators. Operators deploying AI-based planning engines are already reporting success metrics: However, AI alone will not deliver results. Operators need to embed those simulation engines into their capital process and governance. Budget decisions need to be explicitly tied to the outputs, planning teams need to be trained to act on and improve model outputs, and engineering teams must provide feedback, so models consider new or unexpected restrictions. AI is also reengineering and optimizing network operations across multiple areas: In energy management, AI dynamically optimizes energy consumption by managing sleep features and detects anomalies without affecting service quality. In field operations, route optimization and automated scheduling reduce idle time and unnecessary dispatches. In maintenance, predictive models shift operators from reactive repairs to proactive interventions on critical assets. Combined, AI-driven operational use cases can reduce total network opex by 15 to 30 percent. To maximize value, operators need to redesign current workflows around human–AI collaboration, automating workflow steps, eliminating redundant handoffs, coordination efforts, and inefficiencies. Issue resolution and self-healing capabilities are emerging as one of the most widespread AI applications in network operations. Operators are deploying AI across the entire “issue management journey.” For example: Advanced anomaly detection and CX monitoring models identify potential faults early, even before customers are aware of a problem. Smart co-pilots and root-cause models analyze historical incidents and equipment documentation to recommend remediation steps, while dynamic matchmaking systems assign tickets to engineers with the most relevant expertise. Critical change agents automatically identify planned updates with disruption risks and design fallback or remediation plans to prevent major disruptions. At scale, these capabilities have enabled operators to achieve 30 to 70 percent fewer troubleshooting tickets, leading to 55 to 80 percent reductions in network operations center costs, and 30 to 40 percent faster mean time to repair, alongside measurable improvements in customer experience. As operators move beyond pilots, four AI-enabled shifts are emerging: Access to AI doesn’t separate leaders from laggards; instead, it’s the ability to embed AI as a continuous management capability across the network life cycle. Top performing operators take a business-led approach to AI, prioritizing use cases by value rather than technology, investing early in data foundations and digital twins, building internal capabilities beyond pilots, and partnering selectively with hyperscalers and AI specialists. Crucially, they recognize that AI is evolving at exceptional speed—models, tooling, and use cases are advancing every month. AI is now the primary lever to reset network economics, protect margins in mature markets, enable disciplined growth in emerging markets, and unlock new, differentiated revenue streams. Operators that industrialize AI end-to-end can materially outperform peers on ROIC, EBITDA resilience, and long-term strategic optionality. Gustav Grundin is a McKinsey partner in the Helsinki office; Sebastian Cubela is a partner in the Miami office; and Tomás Lajous is a senior partner in the New York office, where Borja Belda is an associate partner. Matyas Zetek is an associate partner in the Prague office, and Sebastián González is a consultant in the Connecticut office. The authors wish to thank Abhimanyu Harlalka, Cindy Xiao, Eduardo Zamudio, and Manuel Palacios for their contributions to this article. This article was edited by Daniel Eisenberg, an executive editor in the New York office. Never miss an insight. We'll email you when new articles are published on this topic. These cookies allow us to count visits and traffic sources so we can measure and improve the performance of our site and app. They help us to know which pages are the most and least popular and see how visitors move around the site and app. All information these cookies collect is aggregated and therefore anonymous. If you do not allow these cookies we will not know when you have visited our site or app, and will not be able to monitor its performance. These cookies enable the website and app to provide enhanced functionality and personalization. They may be set by us or by third party providers whose services we have added to our pages. If you do not allow these cookies then some or all of these services may not function properly. These cookies may be set through our site or app by our advertising partners. They may be used by those companies to build a profile of your interests and show you relevant adverts on other sites. They do not store directly personal information, but are based on uniquely identifying your browser and internet device. If you do not allow these cookies, you will experience less targeted advertising. These cookies are necessary for the website and app to function and cannot be switched off in our systems. They are usually only set in response to actions made by you which amount to a request for services, such as setting your privacy preferences, logging in or filling in forms. You can set your browser to block or alert you about these cookies, but some parts of the site and app will not then work. These cookies do not store any personally identifiable information.