Future 6G mobile networks will evolve beyond basic connectivity to become platforms that host AI agents capable of automating complex enterprise workflows.
Where 5G emphasised bandwidth and low latency, early standardisation work indicates that the next generation of mobile infrastructure will operate as a distributed intelligence layer, able to interpret business intent and coordinate autonomous actions across devices and systems.
This shift transforms the network from a passive data conduit into an active orchestration layer. A recent draft submitted to the Internet Engineering Task Force (IETF) describes how operators plan to deploy “network AI agents” that interact with on-device and third-party agents to carry out sophisticated tasks. While this architecture promises new levels of automation and efficiency, it also imposes stringent requirements for identity management, security, and cross-domain governance.
6G networks: From connectivity to intent-based operations
The core proposal emerging from the 3GPP R20 research is the notion of intent-based services. Instead of manually configuring technical parameters, users or enterprise systems express a high-level objective—such as coordinating a rescue operation or maintaining connection quality for a critical meeting—and the network determines the optimal way to achieve it.
In a disaster response example, a command centre could issue an intent to “conduct a rescue mission with multiple robots in a specified area.” Network-native AI agents would then take over planning, breaking the high-level objective into discrete tasks like obstacle sensing, route planning for individual robots, and allocating communication and compute resources to the affected zone.
By enabling decision-making closer to the edge, this capability reduces reliance on central cloud systems and accelerates the loop between perception and action. For logistics and field operations, the network itself can handle real-time coordination so human operators can concentrate on strategic choices rather than connectivity details.
Orchestrating cross-silo data with AI agents
This architecture is also valuable in commercial settings where data is isolated across different providers. The IETF draft presents an electric vehicle (EV) charging use case that illustrates secure, cross-domain collaboration.
An EV’s on-device agent could monitor energy prices and consider selling stored battery power back to the grid. Before doing so, it would securely check the vehicle owners’ calendars—hosted by various providers—to confirm travel plans. If the agent detects a planned 900 km trip, it would cancel or delay the sale to ensure the vehicle remains sufficiently charged.
Such automation demands rigorous privacy and access controls. The draft stresses that cross-border data exchanges must prevent unauthorised access and ensure an agent only retrieves data it is explicitly permitted to use. Efficient collaboration protocols therefore need to support multimodal data interactions while preserving data sovereignty and legal compliance.
Service exposure and new revenue streams
For telecom operators and enterprise customers, 6G creates opportunities to monetise more than connectivity. Networks could expose sensing, computing, and AI/ML services directly to third-party applications and corporate systems.
Take autonomous transport: operators have unique, wide-area environmental visibility and distributed AI capabilities. A network-native AI assistant could interpret a vehicle’s intent for “safe navigation” and orchestrate local inference, external data integration, or resource allocation to meet that goal.
Similarly, for business travellers the network could offer personalised service provisioning. If a user has an important online meeting while travelling by train, a 6G network-native AI agent could analyse the route, predict coverage gaps, and pre-provision resources to maintain the required quality of experience. This approach shifts service guarantees from static service-level agreements to dynamic, agent-negotiated assurances.
The governance imperative: Identity and reliability
Introducing autonomous agents into the network fabric also raises significant risks. The IETF draft warns that “security risks (malicious intent, intent misinterpretation) of AI agents are critical.” A poorly authenticated or malicious agent could disrupt operations or expose sensitive enterprise data.
To address these threats, standardisation efforts propose a robust identity framework. Networks must implement secure authentication, authorisation, and lifecycle management specifically for AI agents, separate from human user credentials. This includes verifying the identities of on-device agents, third-party agents, and network agents before permitting interaction.
Reliability assurance is equally important. Because agent decisions can modify network configurations directly, operators should deploy safeguards—such as network digital twins or simulation backstops—to validate decisions before they are enacted. Only actions verified for reliability and safety should be allowed to alter live network states, reducing the chance of cascading failures from autonomous errors.
Preparing for AI agent-native infrastructure in the transition to 6G networks
The transition to 6G will reshape how enterprises engage with telecom providers. Embedding AI agents into the network enables operators to unlock more of their infrastructure’s value, improving operational efficiency and end-user experiences.
Enterprise decision-makers should track the development of agent communication protocols and incorporate agent support into future procurement and integration strategies. Organisations will need to update identity and access management (IAM) policies to distinguish user identities from agent identities and to govern non-human entities acting on behalf of employees.
As the 3GPP R20 study advances, attention will focus on how agents are registered, discovered, and governed. Success will depend on trust frameworks that allow diverse AI-driven devices—from robotics to smart cameras—to collaborate securely without compromising network integrity or privacy.
Want to learn more about AI and big data from industry leaders? Attend the AI & Big Data Expo events in Amsterdam, California, and London to hear from experts across the industry. These events are part of TechEx and are co-located with other leading technology conferences.
This article is published by TechForge Media. Explore other enterprise technology events and webinars organised by the publisher.