Roque Lozano, SVP of network infrastructure for MEA at Nokia.
Artificial intelligence (AI) is expanding the boundaries of what’s possible in business, but behind this is a complex, unseen layer of infrastructure that must rise to meet AI’s demands – starting with the network.
This is according to Roque Lozano, SVP of network infrastructure for MEA at Nokia.
Speaking to ITWeb on the sidelines of AfricaCom 2025, hosted recently in Cape Town, Lozano said AI is ‘a double-edged sword’ in the digital world, but particularly when it comes to networks.
“On the one side, [it] is stretching the network performance requirements like never before. On the other side, it’s one of the biggest enablers we have to improve the performance of such networks – so it’s a challenge and solution at the same time,” he explained.
While Nokia has been using AI and machine learning protocol and processes for some time, said Lozano, the advent of GenAI and AI adds a new perspective in terms of the scale and size of AI implementations across all industries.
“Now the network must be stretched to distribute AI services across the entire network, not only in the summary network, not only in data centre connectivity, but within enterprise and consumer connectivity.” This is a crucial step to be able to successfully implement AI and eventually secure ROI.
Lozano added that another advantage of AI is its ability to generate better and more meaningful content – and serve as an enabler for the consumption of that content by improving the network performance and operation.
Nokia’s role is to engineer AI for networks and networks for AI, he said.
“AI is a strategy that is helping you…without AI globally, we would not be able to have the network we need to deliver AI. We are structuring all our developments, all our portfolio strategy to make sure that on the one side, we use AI for networking performance – mainly on the automation, to build the autonomous network as soon as possible. On the other side is the network for AI, that requires latency, safety, security, scalability, resilience that bring all the physical and economic flows to the stream.”
Local demand for automation
Nokia’s recent MEA Automation Study, conducted by Omdia, saw MEA respondents reporting 16% higher levels of network automation than the global average across network domains.
Additionally, MEA operators anticipate extending this lead and are projecting network automation maturity levels around 21% higher than their global peers in the next three years.
According to Nokia, this evolution is driven by the increasing complexity of modern networks, the rising demand for real-time services, and the need for greater operational efficiency.
At the centre of this transformation is a move toward intelligent automation – enabling proactive, predictive and self-healing networks that not only scale with demand but also reduce human error and operational stress.
Samar Mittal, VP and head of cloud and network services for MEA at Nokia, said: “As Nokia, we are addressing this complexity with an operational framework based on three key functions: sense, think and act. By ‘sensing’ data from across the network, systems can detect anomalies, usage patterns and potential faults in real-time. AI and ML algorithms then ‘think’ by analysing this data using extensive knowledge bases – built from both local and global intelligence – to generate actionable insights.
“Finally, autonomous orchestration tools ‘act’ by applying closed-loop automation to mitigate issues, optimise performance and deliver quality at scale.”
Samar Mittal, VP and head of cloud and network services for MEA at Nokia.
According to Mittal, this approach doesn’t eliminate the need for human oversight; rather, it elevates human capability. In environments where engineers previously spent hours navigating logs or troubleshooting errors, AI now handles the heavy lifting – freeing professionals to focus on strategy, innovation and upskilling.
“The human element becomes more critical as the complexity of AI-driven systems increases, with skilled engineers guiding system learning, validating insights and continuously tuning performance,” he said.
Nokia asserts that the continuous evolution of AI and cloud technologies is reshaping the future of network operations by reducing operational errors, supporting greener operations through energy-efficient automation, and enhancing resilience by managing the complexity introduced by cloud-native architectures, data centre proliferation and multi-vendor ecosystems.
While the Nokia Threat Intelligence Report highlighted the increased use of GenAI to enable faster and more sophisticated attacks, communication service providers are also increasingly applying the same technology to combat these attacks.
With the rise in network-based threats and increasing regulatory scrutiny, AI-driven threat detection and identity access management tools offer a more robust line of defence.
Mittal continued: “These technologies, fuelled by global threat intelligence and adaptive algorithms, aim to deliver what we describe as a ‘zero-trust security network’ — a model in which every access point, device and transaction is verified before being trusted.
“As hyperscalers and digital-first business models gain ground, the demand for intelligent, secure and agile network infrastructure will only intensify.”
Mittal and his colleagues believe AI and cloud orchestration provide the foundation to not only meet these demands, but also enable faster service innovation, better customer outcomes and more sustainable operations.
