This paradigm is now beginning to fade. Software is increasingly determining the functionality and intelligence of networks. From this reality and the rise of AI agents, capable of translating human intentions into technical actions, a new form of infrastructure is emerging: open, programmable, and accessible to far more people than before. Network equipment providers, such as Nokia, are also developing these concepts, as they represent the logical evolution of modern networks.
From a hardware-centric network to a programmable infrastructure
The transition to software-defined networking has laid the groundwork for this change. Network functions are no longer configured primarily through dedicated hardware, but rather through APIs, code, and automation. Network as Code takes this approach a step further, transforming functions such as prioritization, latency optimization, and quality of service into software modules. These can be managed through predefined interfaces, without requiring developers to have in-depth knowledge of the network's technical architecture. This opens up a new area of innovation that extends far beyond telecommunications. Companies in sectors such as logistics, manufacturing, research, and media can integrate network logic without needing specialized expertise.
AI agents as a new layer of intermediation
However, the real breakthrough comes from AI agents, which understand objectives, make decisions, and execute actions, often across multiple technical systems. Crucially, they don't operate with parameters, like conventional automation tools, but rather are intent-based. For example, if a team needs stable video quality to conduct interviews in a remote area, the AI agent translates this requirement into technical measures. It prioritizes available upload bandwidth, analyzes local network conditions, optimizes the connection through the nearest radio cells, and dynamically adapts if weather conditions or network load change. This gives rise to a new form of collaboration between people and infrastructure, where technical complexity takes a back seat.
Innovation is becoming widespread
With this evolution, who can drive innovation also changes. Until now, only those with in-depth knowledge of telecommunications could design network logic. Today, developers at startups, logistics companies with connected fleets, Industry 4.0 teams, field researchers, and mobility providers can directly access network intelligence. Innovation is thus more widely distributed, similar to other moments in history when technology opened access to knowledge or resources. The printing press is probably the best-known example, but cloud computing also falls into this category. Programmable networks could play a similar role in digital infrastructure.
The future is already visible
Numerous practical examples demonstrate that this evolution is real. Orange uses AI-based network functions to improve the safety of field personnel. Deutsche Telekom enables high-precision drone operations that rely on dynamic network optimization. The operator Elisa, together with Elmo Cars, is testing remote driving scenarios where latency and reliability must be orchestrated in real time. Meanwhile, WaveXD integrates Network as Code directly into developer platforms, allowing teams to use network functions as easily as software libraries. The democratization of infrastructure is not a future promise: it's already happening.
To what extent can a critical network be opened up?
Despite all the possibilities, it's clear that telecommunications networks are also critical infrastructure. Security, regulatory compliance, and service quality must be guaranteed at all times. The central challenge lies in designing systems that enable maximum innovation without losing control. The outlines of a solution are already beginning to emerge: network operators retain responsibility for security and performance, while AI agents and programmable interfaces create controlled and secure access points. Companies and developers gain new freedoms, but within a clearly defined framework. It's a delicate balance that, nevertheless, opens up enormous opportunities, both socially and economically.
A new group of innovators
We are at a turning point. Networks are no longer just used, but actively designed: by many, not just a few. AI agents and Network as Code are opening the door to a world where innovation is more widely distributed, network logic is embedded in applications, and businesses can experiment faster. For the first time, people from diverse sectors have access to infrastructure intelligence. The question "Who gets to innovate?" thus finds a new answer: (almost) everyone who wants to solve a problem.
Source: Nokia
