Mobile core in 2026 and beyond: the role of AI, agents and 5G-Advanced

On 26 November, I’ll be attending the 5G Core Summit in Bangkok – the fourth time in a row I’ve attended the event. As much as RAN innovations touch users directly, the mobile core is critical for enabling revenue-generating service innovation. And, because any given core node potentially touches millions of users, the performance and resilience profile of core assets is essential. Of course, the value of attending the same event multiple years in a row is seeing what’s changed and how the industry is progressing. Against that backdrop, here’s what I am looking out for.

5G standalone (5G SA) and 5G-Advanced

After a slow start, operators are accelerating 5G SA deployments. As of September 2025, 78 operators had launched 5G SA in 42 countries and 96 had expressed intent to launch. This means a total of 174 operators in 72 countries. Meanwhile, 5G-Advanced is being deployed in pioneer markets such as China, the US, and a few markets in the GCC and Asia Pacific. Enhancements in network performance, management and efficiency are core 5G-Advanced value propositions, enabling a faster shift to more autonomous networks, which in turn enables embedded network intelligence. Equally interesting are IoT-related innovations such as added positioning accuracy, RedCap and tighter non-terrestrial network (NTN) integration. 

For the mobile core, this means new service types (including AI services) and new device types (including AI devices) to manage. Beyond an obvious focus on extending 5G’s push into B2B services, enhanced network and AI capabilities (highly interrelated) open the door to new, immersive communications services. AI calling is a notable example, as AI allows operators and vendors to take the New Calling service to a higher level of engagement across multiple features such as real-time translation, fun calling and voice-driven avatars, and interactive menus powered by personal digital agents. The mobile core will need to adapt to these new device and service types, addressing added capacity demands as well as different service types (high versus low bandwidth, high versus low latency, voice versus video versus data) in a deterministic manner. 

Artificial intelligence 

AI for networks is a growing focus for operators (and will be a key topic at the 5G Core Summit), with use cases spanning network optimisation, troubleshooting, security support, energy efficiency, and network management. In many ways, this can be seen as an extension of the 5G SA and 5G-Advanced dynamics. The speeds and uplink capacity of 5G-Advanced support AI traffic loads. Low-cost IoT, in turn, sets the stage for AI data collection efforts. The result is a new set of devices carrying diverse service and traffic types. Again, traffic requirements and determinism will come into play – particularly as AI becomes part of mission-critical business processes. And, if AI is touching high-value data, security will take on a greater importance. 

AI’s impacts on the mobile core could be particularly disruptive in terms of distribution. To deal with security, latency and resilience demands, many AI use cases could require distribution of core assets out to the network edge. Of course, combining a broad set of distributed nodes with new traffic demands implies a new set of O&M requirements for the mobile core as well. 

Agentic AI

The promise of agentic AI is straightforward; agents that plan, reason and act independently can coordinate complex tasks and processes, even engaging other agents where end-to-end, cross-domain orchestration is required. In practice, this represents an extension of today’s automation tools – an extension necessary for addressing the added network and service complexity ushered in by 5G SA/5G-Advanced and AI workloads. Core operations and maintenance systems have embraced intelligence and AI capabilities to deal with complexity and efficiency demands. The agentic core takes this further – hence, AI-native core. GSMA Intelligence’s annual global survey of operators shows that two in three operators have either started to deploy agentic AI technology in their core networks or are currently testing it. That means core networks are a major focus for agentic AI. I look forward to seeing discussions on this at the 5G Core Summit.

Monetisation 

The 5G monetisation imperative can only grow, as operators invest in 5G SA and 5G-Advanced. An increasing number of operators are focusing on experience-based consumer propositions (i.e. monetising user experience through differentiated services, including experience/service assurance). Operators in China, AIS Thailand, Singtel in Singapore, Elisa in Finland and some of the Middle Eastern operators are notable examples. It will be interesting to see new examples at the 5G Core Summit. Core network intelligence is key to achieving experience-based monetisation; this involves leveraging AI, 5G SA and 5G-Advanced to optimise network performance and deliver the desired user experience while providing effective service assurance. 

B2B will be a focus too. GSMA Intelligence research, based on a new global survey of 5,320 enterprises across 10 vertical sectors and 32 countries, shows that enterprises will spend 10% of their revenues on digital transformation during 2025–2030, with AI and 5G the top areas of investment. This represents an opportunity for the mobile industry to help enterprises advance their digital transformation while growing B2B revenues. 

Putting it all together

While I’ll be looking to see how these dynamics are addressed at the 5G Core Summit, it would be a mistake to think of them as standalone themes. They are all linked, and those linkages drive other core network priorities I’ll be watching out for. AI, for example, empowers 5G-Advanced, enabling new services and use cases. Core networks will play a critical role in making that happen but will also need to make use of AI if they want to do it efficiently and resiliently. The interplay between 5G-Advanced and AI represents the near-term focus for operators, extended by the power of agents in support of network and service automation. However, it also points to the central role of openness in core network evolution. If new services and devices are to communicate their requirements, and agents are to coordinate among themselves, closed systems will only hold back progress. 

If you’ll be joining us in Bangkok, reach out – it would be great to connect. If you’re in the region but don’t have it on your agenda yet, consider making the trip. I’m expecting it to be an incredibly valuable event.