5G-AI integration is reshaping technology and digital sovereignty

Fifth-generation networks, and the 6th generation currently under development and set for deployment around 2030, enable the uptake of advanced AI applications that require processing larger data volumes, higher data speeds, and real-time communication across diverse user experiences.

Such interactions demand the rapid transmission of data between device and network: consistently low latency (delay) and high uplink throughput. Standard 5G networks significantly reduce latency compared to 4G, typically achieving around 10ms and theoretically reaching as low as 1ms under optimal conditions (compared to 4G's 50-100ms). This latency improvement increases processing speed, including AI inference and data transmission, thereby enhancing responsiveness by 50–60%.

But current 5G architectures predominantly prioritize downlink (download) capacity, a legacy of earlier networks optimized for consumer activities like video streaming and gaming. Consequently, uplink bandwidth—essential for real-time, AI-driven use cases—often lags.

Therefore, fully realizing the consistently ultra-low latency and robust uplink capacity necessary for advanced AI applications will require further evolution of 5G networks into AI-centric infrastructures, integrating AI-driven management and operations directly into their architecture.

From infrastructure to value creation

Current networks won’t be able to keep up. Some conventional 5G networks are already stretched by “network traffic waste,” a term that refers to inefficient or unnecessary use of network resources caused by features such as auto play, infinite scroll, and pre-fetching of content – features incentivized by the business model of social media platforms. Vodafone’s recent policy statement refer to this as a digital “tragedy of the commons.”

While "Responsible use of Networks" is vital for addressing energy efficiency, the reality is that data traffic will continue to grow exponentially. This strain exposes a fundamental mismatch between traditional network architectures and emerging data-intensive AI applications. It makes a shift towards AI-centric network operations and management not just sensible, but indispensable if we are to manage rapidly growing data efficiently and enable highly flexible, real-time, high-bandwidth use cases. Such 5G-AI integration offers telecom operators a unique opportunity to accelerate 5G adoption by unlocking new use cases across various sectors.

However, this shift requires substantial investments in network infrastructure, in addition to already high investments in computing infrastructure and cloud services. As a result, while the AI-driven digital transformation offers tremendous productivity potential, it is also far more capital-intensive than the initial Internet revolution.

EU lags in network infrastructure buildout

As of 2024, the European Union had made substantial strides in expanding 5G coverage, which now reaches approximately 87% of the EU’s population. However, it still lags behind the United States (98%) and South Korea (99%), both of which have achieved near-universal 5G availability. Despite the growing infrastructure, the actual 5G connection ratio in Europe remains relatively low at around 30%, compared to higher adoption rates in these leading markets.

Looking ahead, the EU is projected to reach 80% of mobile connections on 5G by 2030, indicating a slow but steady transition. This delay is largely attributed to the fragmented European telecom market, with varying regulatory environments and investment levels across member states. Moreover, many end-user applications are still adequately supported by 4G (e.g., video streaming, mobile gaming, and social media, as well as messaging and video calling, e-mail, web browsing, and cloud storage). All of this means that compelling consumer use cases for 5G have yet to materialize at scale.

Although Europe has achieved broad 5G coverage, its deployment of 5G Standalone (SA)—the architectural foundation needed to support AI-centric capabilities such as network slicing and ultra-low latency—remains minimal, with just 2% SA availability in Q4 2024, compared to 24% in the U.S., 52% in India, and 80% in China. This lag presents both a challenge and a strategic opening: by accelerating investment in 5G SA and embedding AI into network management and operations, telecom operators could shift from being a technology taker to a standard-setter in the AI-driven digital economy. This underscores the urgency of adopting an Infrastructure First approach. Rather than waiting for compelling applications to drive demand, the focus must shift to building AI-ready infrastructure proactively.

A paradigm shift to intelligent infrastructure

These capabilities signify a paradigm shift from static, hardware-centric networks to intelligent, software-driven infrastructure tailored for AI ecosystems. By adopting this transformation, telecom operators transcend their role as connectivity providers to become strategic enablers of AI innovation across industries.

Achieving this shift demands sustained investment in three pillars: a unified infrastructure layer that integrates edge-to-core networks to support low-latency AI data flows; an intelligent control layer that dynamically manages resources and resolves legacy 5G asymmetries, especially uplink limitations; and an autonomous operations layer that enables predictive optimization and real-time self-healing to ensure resilient, SLA-compliant performance for AI-driven applications. Through platforms like GSMA Open Gateway, operators can expose AI-optimized network capabilities—such as latency guarantees, bandwidth slicing, or reliability SLAs—to third parties, thereby generating new revenue streams (e.g., “network-as-a-service” models tailored to sector needs), establishing cross-sector ecosystems (e.g., partnerships with smart cities, manufacturers, healthcare providers), and driving policy innovation (e.g., developing secure, scalable access frameworks that balance open innovation with digital privacy and security regulations).

Unlocking this value, however, requires more than technical upgrades. It calls for systemic transformation. This evolution positions telecom networks as critical infrastructure for the AI economy, driving efficiency, safety, and economic growth while requiring collaboration between governments, industries, and regulators to standardize and scale. Yet despite this strategic evolution in the role of telecom networks, regional disparities in deployment and adoption remain significant.

A transformative leap

Looking ahead, 6G will mark a transformative leap by integrating AI directly into network architecture, evolving networks from connectivity platforms into intelligent infrastructures capable of proactively managing themselves. Such “AI-native” design will enable networks to predict and autonomously adapt to demands, faults, and security threats, drastically enhancing efficiency and sustainability. Novel capabilities such as joint communication and sensing, semantic communication, and distributed AI will enable entirely new services, from hyper-accurate positioning to meaningful data exchange, while safeguarding privacy and minimizing energy consumption. This convergence means 6G will not just support digital economies—it will create them, necessitating forward-looking policies that balance innovation, regulation, and digital sovereignty.

For an economy to fully realize AI as a new driver of productivity, telecom operators must integrate AI into the operations and management of their 5G networks. This 5G-AI integration is critical—not only to deliver the connectivity and bandwidth needed for exponential data growth, but also to achieve the ultra-low latency required by advanced AI applications.