Hello,

This week, operators in Asia, Europe, and Africa deployed autonomous AI systems at scale. Same technology. Same timeline. Different continents.

In today's edition:

Let’s dive in.

Telecom's autonomous network era arrived this month. Multiple operators deployed agent-based AI systems that make decisions without human intervention.

Docomo launched an agent-based platform on February 4, 2026. Built on AWS Bedrock AgentCore, the system analyses data from over one million network devices in real time. Multiple AI agents work together, orchestrated by network topology graphs. They identify anomalies, isolate root causes, and suggest fixes. Fault resolution times dropped by more than 60 per cent.​

Far EasTone Telecom turned this architecture into operational reality, announced at MWC 2026 on February 23. Nearly 60 per cent of its NOC operations now run AI-assisted. The system executes about 10,500 operational tasks per month. AI agents handle alarm correlation and root cause analysis in seconds. They support nearly 7,000 monthly operational queries with 16-second average response. Most maintenance actions complete within one minute. The shift reduced human error and accelerated recovery times whilst allowing engineers to focus on higher value work.​

Vodafone is applying Microsoft's proven blueprint to transport infrastructure and field-force management, also announced February 23. The collaboration combines Vodafone's network expertise with Microsoft Foundry and the Network Operations Automation framework. Microsoft runs autonomous agents across its global Azure transport network, where AI autonomously manages more than 65 per cent of fibre-break field dispatches. Time to repair improved by up to 25 per cent. Root-cause analysis accelerated by 80 per cent.​

The pattern holds across regions. AT&T, T-Mobile, Telefónica, and MEO adopted Microsoft Foundry as their agentic AI blueprint. Kenmei announced collaboration with Microsoft to help operators reach autonomous networks by combining Kenmei's telecom intelligence with Azure and Microsoft Fabric. Already in use at Telefónica and Etisalat, the platform reduces manual effort and speeds decision-making.​

Three forces converge. First, operators need measurable returns from AI investments. Gartner forecast worldwide AI spending to total $2.52 trillion in 2026, published January 18. That's 44 per cent year-over-year growth. Gartner noted AI sits in the Trough of Disillusionment throughout 2026. Enterprises prioritise proven outcomes over speculative potential. They buy AI from incumbent software providers rather than as moonshot projects.​

Second, network complexity demands automation. Modern networks span multiple vendors, technologies, and domains. Human teams cannot monitor millions of data points in real time. Agentic AI fills this gap.

Third, the shift from automation to autonomy changes economics, published February 22. McKinsey research shows customers are up to five times more likely to churn after poor network moments. Acting on network experience signals instantly delivers the most value. Operators can't wait for batch processing or scheduled reports.​

What this means for you: Agentic AI deployment requires architectural choices now. The successful operators integrated Microsoft Foundry, AWS Bedrock, or equivalent platforms before deploying agents. They didn't build from scratch. Platform selection determines which AI capabilities you can deploy in 12 months.

Ericsson emerged as clear leader in ABI Research's 5G RAN Automation Platform competitive ranking, published February 11. The analysis compared eight global vendors on innovation and implementation. Ericsson's Intelligent Automation Platform scored highest overall and led the implementation ranking.​

The ranking reflects commercial reality. EIAP runs on six live networks at different deployment phases. Customers include AT&T, Swisscom, Telstra, and Vodafone across North America, Europe, and Asia-Pacific. The platform proved multi-vendor interoperability and integration in production environments.​

ABI Research highlighted Ericsson's rApp ecosystem. Nearly 90 member companies participate, including more than 20 CSPs. They made approximately 90 rApps available. Ericsson contributes over 25 rApps. The report noted comprehensive developer support, including GenAI assistance for pro-code and low-code developers. The platform manages all RAN types: purpose-built, Cloud RAN, Ericsson systems, and third-party equipment.​

Swisscom uses EIAP for AI management of multi-vendor, multi-tech 4G/5G RAN systems, announced February 22. The platform includes Ericsson's non-real time RAN Intelligent Controller and an open SDK for operators to deploy rApps that automate RAN functions. Swisscom is exploring AI-driven automation potential.​

Spain's MasOrange accelerated O-RAN deployment with Ericsson, reported December 11. One-third of its 5G sites are now O-RAN-ready. The operator integrated EIAP into network management. rApps automate network planning, performance analysis, anomaly detection, and energy management. Two key applications run live: Ericsson's Cell Anomaly Detector uses AI to identify issues with coverage, latency, and interference. Future Connections' Nix RAN Energy Saver reduces consumption whilst maintaining service quality.​

NVIDIA released its 2026 State of AI in Telecommunications survey on February 18. The survey polled over 1,000 telecom professionals globally. Results show AI became the core growth engine for telecom operations, networks, and services. Operators invest in efficiency gains and untapped revenue. They plan for growing AI traffic and high-value services as they build distributed AI computing infrastructure.​

Strategic Analysis

Platform selection determines automation speed. Operators choosing established platforms with proven ecosystems deploy faster than those building custom solutions. The rApp model accelerates deployment because operators and third parties develop applications independently.

Multi-vendor interoperability matters more than vendor preference. Networks include equipment from multiple suppliers. Automation platforms must manage all of it. Operators can't wait for vendor lock-in to resolve itself.

The competitive ranking signals market maturity. When research firms publish structured vendor assessments, the technology moved beyond early adoption. CSPs now compare features, ecosystems, and commercial terms rather than questioning whether to deploy.

What this means for you: If you're planning RAN automation, evaluate the rApp ecosystem size and quality. A platform with 25 rApps limits your automation scope compared to one with 90 rApps from diverse developers. Ecosystem breadth determines how fast you can automate new workflows without building applications yourself.

MTN Ghana and Huawei completed the world's first large-scale deployment of the Alpha Antenna, announced February 14. This represents substantial network performance enhancement. The transition to intelligent, AI-driven operation sets a new benchmark for autonomous driving networks in Africa and globally.​

Post-deployment tests recorded regional traffic rising by 6.8 per cent. Operations and maintenance efficiency improved by a factor of 30x. MTN Ghana stated the Alpha Antenna represents a pivotal step toward full digital transformation of network infrastructure. It reduces operational complexity and costs whilst empowering the network with real-time retrieval and rapid-response capabilities. The collaboration with Huawei laid a foundation for future AI-driven intelligent networks.​

The African Development Bank, UNDP, and private partners launched the AI 10 Billion Initiative at the Nairobi AI Forum on February 23. The partnership seeks to mobilise up to $10 billion by 2035. Resources will unlock up to 40 million new jobs across the continent through targeted investments building AI foundations and catalysing adoption. Focus areas include entrepreneurship, regional data infrastructure, policy frameworks, and skills development.​

The initiative follows a June 2025 Bank Group report outlining a three-phase roadmap towards AI readiness. Five interlinked enablers guide the work: data, compute, skills, trust, and capital. The Bank Group claims AI could generate up to $1 trillion in additional GDP by 2035 if developed and deployed inclusively. This reflects Africa's demographic advantage, growing digital capacity, and sectoral reform.​

AI-powered radio networks emerge as the next frontier for African operators, reported January 9. Rising mobile adoption, higher data consumption, and demand for advanced digital services increase pressure to deliver better network performance. AI-RAN embeds artificial intelligence into network management, enabling operators to optimise resources, improve service quality, and engage users proactively.​

Vodacom collaborates with Nvidia and Nokia to build AI-enabled network management platforms using machine learning for operational decision-making and performance optimisation. MTN partnered with Rakuten Symphony, Accenture, and Tech Mahindra on O-RAN proof-of-concept trials in South Africa, Nigeria, and Liberia, laying groundwork for AI-assisted RAN capabilities.​

What this means for you: Africa's AI infrastructure investments create equipment and platform opportunities now. The $10 billion initiative and operator deployments signal coordinated momentum rather than isolated pilots. If you're planning African market entry, focus on solutions proven in capital-constrained markets. Cost reduction matters more than feature richness. The 30x efficiency improvement in Ghana demonstrates this priority.

Latin American telcos shifted strategy for 2026, reported February 9. After a cautious 2025 focused on stabilising operations and improving efficiency, operators now bet on 5G monetisation, RCS messaging, satellite connectivity, and AI. Last year operators prioritised consolidating existing capabilities, reviewing strategies, and preparing for the next technological cycle. 2026 brings technology consolidation.​

Artificial intelligence took central place in strategic conversations despite not yet functioning as a clear competitive differentiator. Operators perceive AI as a true game-changer, especially for customer service. Concrete use cases focus on process automation, incident resolution, customer behaviour analysis, and optimising user experience across service channels. AI emerged as key tool for improving operational efficiency, reducing friction, and scaling interactions intelligently.​

Enterprise AI in Latin America enters consolidation phase, reported January 30. Greater business maturity and progress of AI agents position 2026 for concrete returns in productivity and operational optimisation. The shift from experimental pilots to proven deployments changed investment decisions.​

The Asia-Pacific region provides context, published January 22. Operators accelerate the shift toward next-generation connectivity and responsible AI. Enterprises gain momentum through AI and platformisation, supported by TrueBusiness's AI Hub, SK Telecom's national teacher model, and KDDI's BSS modernisation. Cross-border AI infrastructure led by NTT, Chunghwa Telecom, and Edgecore connects ecosystem enablers.​

IBM's APAC AI Outlook 2026, released January 7, shows 64 per cent of organisations redirecting AI investments toward core business functions where impact on customer value and top-line growth is greatest. By 2026, 95 per cent of global executives expect generative AI initiatives to be at least partially self-funded, reflecting widening revenue pools AI creates across industries. The mindset shifted from cost savings to value creation.​

What this means for you: Latin American operators moved past analysis paralysis. If you sell AI platforms or applications to this region, focus on customer service and operational efficiency use cases with clear ROI. Operators won't fund speculative projects. They want automation that reduces manual effort, speeds decision-making, and scales intelligently. Proof of concept matters less than production deployment timelines and measurable returns.