AfricaCom 2025: Africa’s Transition to AI-Native Telecommunications

During the Operations Transformation Forum, held as part of AfricaCom 2025, industry leaders outlined an ambitious vision for the future of Africa's telecommunications industry. An announcement outlining a new framework for the use of Artificial Intelligence (AI) technology in the management, creation, and operational processes associated with the network stated that the next step for operators is to develop autonomous networks. Key points to note include:

• The use of AI in operations will result in predictive maintenance and the optimization of network traffic via the use of AI agents and digital twins, thereby superseding the current Operational and Maintenance (O&M) model of rule-based systems in favor of a more intent-driven model of automation.
• Cloud and data integration via a carrier-grade stack solution has enabled operators, such as Safaricom and MTN, to deploy workloads with AI capabilities in controlled environments (as opposed to public cloud environments) while maintaining local compliance requirements.
• There will be various uses of autonomous networks in the near future, including using the TM Forum’s Open Digital Architecture (ODA) Framework and intent-based Application Programming Interfaces (APIs) for trial implementations of autonomous networks by several major African operators.

Though there are many examples of the potential for AI-enabled financial services, cloud-based education, precision agriculture, and platform-based value creation through AI-native networks, most operators in Africa are still in the first steps of building the technical and organizational capabilities and foundations necessary for these scenarios. The ambitions outlined here are representative not only of the African market but of a global phenomenon, which suggests that before any AI-enabled service can be commercially powered on a sustainable basis, the operators must first have consolidated their capabilities in terms of cloud, data, and automation technologies.

The concept of autonomous and self-healing networks enables operators in Africa to cut down on Operational Expenditure (OPEX), enhance the resilience of their networks, and develop new monetization channels. In addition, there is an immediate opportunity for carriers to transition away from the legacy operating model and build out the infrastructure necessary to operate with AI. Without this foundation, the goals of intent-based orchestration will remain as solely aspirations.

Transitioning to an AI-native operations model will not be an easy journey for many telcos in Africa. While some, such as MTN South Africa, Vodacom, and Safaricom, are already operating their networks with a combination of cloud-native principles and Distributed Edge Computing (DEC)—an architecture that places compute and storage resources closer to the user to improve latency, resilience, and real-time processing—only a few have fully deployed cloud-native core networks or Dynamic Spectrum Sharing (DSS) at scale. DSS enables operators to allocate the same spectrum band to multiple radio technologies (such as 4G and 5G) dynamically, allowing more efficient use of limited spectrum resources. While not unique to African operators, fragmented Operations Support System (OSS)/Business Support System (BSS) stacks, as well as legacy systems, are hindering the ability of many service providers in Africa to integrate fully with their networks and gain the maximum benefits from AI-driven automation, challenges that cannot be solved overnight.

Many telcos operating in Africa need to upgrade their existing networks and infrastructure before AI technologies can be adopted by the wider market. These updates will need to address several challenges such as upgrading older network equipment, building out infrastructure to support edge computing, creating new core networks that will be used for 5G-capable architectures, and establishing low-latency communications capabilities across all levels of the network. If these foundational building blocks for modernization do not exist, then telcos cannot scale beyond trial deployments and on a day-to-day basis, perform AI-based orchestration and healing automatically at scale, and use automated network control through intent. Thus, as telcos adopt new technologies, there needs to be a phased approach to rolling them out, balancing investments made into new systems with continued daily operation of current networks so that during a transitional period, performance and reliability of networks are not compromised.

At the same time, there is increasing demand pressure. Huawei’s AfricaCom newsletter states that by the end of 2024, Africa's mobile subscriber base grew to approximately 710 million (almost twice the size of 2019) with average smartphone use escalating to almost 4 hours per day. Additionally, there has been a six-fold increase in download speeds over the past 6 months; therefore, there are now greater expectations for consistent, high-performing services. The increase in demand for low-priced, subscription-style data services has also increased the importance attached to automation. As the bulk of the existing infrastructure has not kept up with this growing need, there is currently a gap between what companies want to do and what they can actually achieve.

The most effective way for African mobile operators to implement AI in their networks is not to begin with automation, but rather with establishing the operational foundation that will enable automation. The use of AI should be viewed as a step-by-step approach toward building an operational backbone (OSS/BSS) for the operator to ultimately automate large parts of the network. Without a solid operational base layer, it is unlikely that mobile operators will have the ability to scale AI solutions across their networks; instead, they would continue to use AI for niche applications at best.

The fastest means of accomplishing this is to provide an integrated environment by reducing fragmentation between the OSS and BSS. Instead of immediately investing in autonomous functions, operators should focus on developing common data models, integrating data in real time, and creating standard telemetry formats; then they can develop consistent views across all operational, service, and customer operations. Together, these actions will create a single operational foundation from which AI applications can be reused and expanded efficiently.

To be ready to leverage AI, operators will have to meet requirements needed for an automated operational environment. They will need to provide proper access to and governance of their data, and create workflows that describe how they operate and build secure control point interfaces. The first set of AI applications deployed by African operators will likely be limited to decision support, anomaly detection, and performance enhancement, with human supervision of these applications remaining while they build trust and experience with AI-based decision-making.

Future advancements of AI-native solutions in the African market will evolve incrementally instead of rapidly. In the coming years, AI-native solutions will improve through experiential improvements with regard to observability, predictive operation, and selective closed-loop assurance, which are impacted by challenges such as capital constraints, outdated systems, and skills shortages. Broad end-to-end autonomy will remain a longer-term objective. Operators should consider the benefits of an OSS/BSS platform focused on AI that will enable operators to create a strong foundational future capability while enhancing current operations. A platform-first approach to operations provides consistency and allows for continued and organic development of advanced automation in response to the organization's ability to implement these systems, as well as changing market conditions.