Juniper Research has published a new study predicting that Artificial Intelligence (AI) will reduce the average operator revenue leakage per 5G roaming connection from $1.72 down to $1.20 by 2024. Revenue leakage, as defined by the telecommunications market experts, is the value of services provided by operators but for which they're unable to monetise.

The firm foresees AI-based segmentation as the key driver of this mitigating effect on revenue leakage. These AI-enabled solutions will allow more efficient allocation and pricing of resources, particularly for data-centric users. The represented reduction in revenue leakage will benefit operators' bottom line and reflects the higher Quality of Service (QoS) achievable with 5G networks that utilise standalone technology, compared to those relying upon 4G infrastructure.

AI-based segmentation methods will utilise machine-learning models to detect and differentiate traffic types in real-time, making it easier for operators to identify and monetise emerging roaming services. As explained by Alex Webb, the author of the study, "AI-based segmentation will differentiate enterprise traffic by use case, enabling premium billing of mission-critical 5G standalone connections, thus reducing revenue leakage."

Juniper Research’s report emphasises the necessity for operators to implement AI segmentation tools to reduce the financial drain from 5G roaming on standalone networks. These networks offer higher throughput and lower latency than their non-standalone counterparts, and this quality difference must be reflected in pricing strategies. By separating standalone from non-standalone roaming traffic, operators can create individual pricing strategies for each type of network, as appropriate pricing will also reflect QoS.

Telecommunications companies, according to the study, must utilise these AI technologies to identify enterprise traffic suitable for use cases for dedicated network slices. Properly implemented, this will optimise network resource distribution and further reduce revenue leakage. The implementation of AI to differentiate and adequately charge for the segments of 5G roaming traffic can trim potential revenue leakage to a projected $118 million by 2024.