日本語
This year’s Mobile World Congress Barcelona was the first in recent memory where it didn’t feel like telecom was chasing what’s next. Rather, it marked a distinct shift toward confident pragmatism, providing an important moment to refine focus on real deployments, revenue generation, and achievable goals, versus futuristic concepts and speculation.
The spotlight was on monetization and growth strategies built on technologies that made headlines at past MWC events.
It’s a sign that telecom is maturing along the 5G journey, with a growing emphasis on learning and collaboration among hyperscalers, system integrators, and enterprises alike.
While AI remained omnipresent, discussions on the show floor, in booths, and during sessions, were grounded not in hype, but in reality, reflecting lessons from early rollouts.
Of course, confidence doesn’t mean certainty. Geopolitics was ever present in the background, with supply chain concerns and trade dynamics casting an unignorable shadow even as progress remained evident.
Here are my top takeaways from another great MWC event.
5G Turns a Corner
Everywhere at MWC were signs that 5G’s ecosystem is maturing as revenue powered by 5G standalone (SA) deployments and 5G-Advanced capabilities begins to materialize.
T-Mobile and Telefónica touted revenue-generating 5G network slicing commercial offerings defined by Quality on Demand (QoD) for emergency services and mission-critical communications.
A full RedCap (Reduced Capability 5G) ecosystem is ready to go, with increased traction demonstrated by Hyundai Motor Company showcasing an end-to-end RedCap trial over a private 5G network, and Sequans, TCL, Pegatron, and Samsung showcasing modules, devices, and services.
Private networks and IoT see growth, with major enterprises showcasing real, yet small-scale deployments. We’re keeping an eye on how the fragmented and complicated supplier ecosystem may hinder larger implementations.
An ancillary theme this year, Fixed Wireless Access (FWA) solutions focused on optimizing in-home quality of experience and coverage, and advanced customer premises equipment (CPE) to support Wi-Fi 7, 4X4 MIMO, and mesh networks. MediaTek showcased a range of solutions for major partner carriers. ZTE highlighted a 5G-Advanced/Wi-Fi 7 FWA CPE with integrated AI for bandwidth optimization, quality of service (QoS) management, and signal tracking. All these solutions were part of an ‘Own the Home’ theme, seen as rich in untapped revenue potential.
ZTE’s 5G-Advanced/Wi-Fi 7 FWA CPE with integrated AI
Read our new FWA eBook for insights into the latest developments.
Operators are ready to enable 5G SA use cases, including QoD, network-as-a-service, and location capabilities. In particular, 5G SA network APIs are seen as a real value multiplier, with over 30 companies partnering in demos at the GSMA Open Gateway Zone.
5G-enabled drones and the growing low altitude economy were a significant theme at MWC with several major operators and vendors showcasing innovative drone technologies and services leveraging 5G networks. Telefónica and Nokia showcased “Open Gateway 5G Drones” for safe and efficient autonomous drone operations and an innovative drone as a service (DaaS) model, while Vodafone highlighted drones for maintenance and inspection.
Telefónica and Nokia’s “Open Gateway 5G Drones”
Read our new blog highlighting the role of the 5G Core in unlocking the low altitude drone economy.
Telecom Leading AI Adoption
Most telcos have been experimenting with AI for years and provided seemingly endless examples of how it is providing real efficiency gains in network operations. No, we are not yet at full AI-driven automation, but the groundwork is laid for future AI-driven operations.
Networks for AI
While early AI progress focused on developments like chatbots, we are now seeing the network as a primary beneficiary of innovation with operators and network gear suppliers leveraging AI as a monetization tool. For instance, e& UAE Chief Executive Officer, Masood Mahmood shared insights on a multi-year journey to use and monetize AI in networks.
Operators want to monetize and optimize networks for AI application traffic, sovereign AI, and to support multi-tenant AI for enterprises where an operator can become a central hub or orchestrator to manage diverse AI models and applications, facilitating seamless integration and adoption that primarily benefits SMEs. On this front, Rakuten Mobile discussed moving from using AI for internal efficiencies to offering AI-driven services to enterprises, essentially becoming an AI hub for customers.
Read our new eBook on testing Networks built for and enhanced with AI.
AI Applied to Radio Access Networks (RAN)
AI-for-RAN (using AI to improve RAN efficiency) and AI-and-RAN (sharing GPU resources between AI applications and radio processing) were both main themes at this year’s show.
AI-driven optimizations for radio networks are already proving valuable and seem on track for widescale implementation. Numerous vendors showed off demos of capabilities that provided enhancements in energy efficiency, optimization of MU-MIMO, and uplink signal processing. It was also good to hear growing practical consensus that AI may not be cost-effective at all RAN layers (such as Layer 1).
The concept of AI and RAN sharing compute platforms remains an open and longer-term debate. Yes, it works in private or low-utilization environments but isn’t yet viable for dense public networks like urban macro deployments. A lot of feasibility studies and testing are required to determine whether it makes sense technically and commercially for widespread adoption.
Generative AI: Where We Are Now
While GenAI and the use of assistants was highly visible, it felt like a known quantity already part of established commercial strategies. Agentic AI, that can act independently, iteratively improve its performance, and autonomously solve complex, multi-step problems, was this year’s hot topic. Discussions centered around governance, human oversight, and gradual evolution toward more autonomous AI agents.
The demos I saw felt like a glimpse into the future of telecom networks with GenAI assistants evolving alongside smaller, more specific AI models into agents, all wrapped in automation. And it was great to hear that operators are evolving their AI thinking with governance and humans-in-the-loop at the heart of their adoption strategies.
On a related note, Spirent showcased its new AI Agent for 5G Core network testing, demonstrating how it can support human users to generate and execute complex test configurations, provide root cause analysis, and offer a level of in-depth guidance.
The Role of Data Centers in AI and Telecom
AI implementations are impossible without scalable data centers. And though this was not a dominant theme at MWC, behind the scenes there was frequent discussion about how data centers must evolve to handle AI workloads and the reality of a future converged network and compute landscape.
Highlights included: HPE’sEX 4000Cray supercomputing for AI cabinet capable of supporting 448 GPUs; SK Telecom’sGPU as a Service (GPUaaS) in cloud and edge data centers; Dell’s PowerSwitch 800GbE AI fabric switches; Senao Networks’ 800G Ethernet switches; Cisco’s GPU scale-out solutions supporting 64 ports of 800G; and Juniper’s AI data center networking showcase.
HPE’sCray supercomputing for AI cabinet
Expect more focus on AI-ready network infrastructure in future events.
Read our HSE markets insight report.
Non-Terrestrial Networks (NTN) and Satellites Shift from Competition to Collaboration
The NTN (space sector) was at MWC in force as a partner and collaborator. There was excitement about NTN’s extensive coverage though the business case still has cost, energy and capacity limitations, and interoperability remains a key challenge.
Overall, I sensed realization that NTN can complement terrestrial networks rather than competing. A big shift in mindset has occurred: The satellite and telecom industries are collaborating and discussing following standards for integration rather than pursuing proprietary approaches. Being complementary is a good business outcome for all parties.
Check out our Space solutions page to learn more.
Augmented Reality (AR): The Next Form-factor Disruptor
AR glasses were showcased extensively and are becoming more practical as designs better prioritize social acceptance and user-friendliness. AR glasses could be the next form-factor disruptor, like the iPhone, and could even challenge the traditional eyewear market. Expect consumer adoption to grow as mass adoption remains a few years out.
Numerous lightweight prototypes were exhibited that tether to a smartphone. They offer ease of adoption and a mass market that’s already conditioned. Many integrate with AI assistants.
For example, NTT QONOQ demoed two models targeting use cases such as navigation, mail notifications, language interpreter, and AI assistants. Techno AI Glasses Pro provided an AI-powered voice assistant, object recognition, real-time information summarization, and real-time navigation. ActiveLook’s AR Smart Glasses showcased a design for lightweight AR glasses optimized for sports and outdoor activities.
NTT QONOQ’s AR Smart Glasses
Quantum Computing: A Future Synergy with AI
Quantum computing was a small but transformative topic. We saw what a rapidly approaching compute and networked world could look like when quantum is combined with AI. There was considerable discussion on how quantum and AI complement each other and will accelerate both adoption and value. A red flag was the need to plan and adopt post-quantum cryptography (PQC) sooner rather than later.
The key focus was on AI-quantum synergies, particularly the role of AI in helping solve quantum computing error correction challenges and how quantum computing will enhance AI models for faster and more complex decision-making. The industry recognizes this will be a major theme in the coming years.
Other takeaways include:
Telefónica and LuxQuanta showcasing how quantum key distribution (QKD) could protect sensitive healthcare data from potential quantum computer attacks, while Telefónica also demonstrated PQC in private 5G networks. Sparkle with Adtran, Arqit, and Intel demonstrated fully automated, on-demand network as a service secured by PQC.
A few suppliers showed quantum repeaters that enable long-distance quantum communication by overcoming the limitations of signal loss in optical fibers.
What comes next after quantum was even discussed, including neuromorphic and synthetic brain computing.
The Pragmatic Future of Telecom at MWC 2025
MWC showed that actual developments in network efficiency, slicing, and AI are making real impacts.
5G is no longer about hype—it’s about monetization and real deployments as the industry converges into a more mature ecosystem. 5G is here to stay, enterprises are engaging, and the technologies are becoming more interconnected.
AI is transforming networks, with both efficiency gains and new revenue models.
NTN, satellites, and hyperscalers are shifting toward collaboration instead of competition.
Quantum computing and AR are emerging and disruptive but will take time to mature.
The industry is moving forward realistically, with a clear focus on near-term results.
It felt promising to see the emerging fruits of the industry’s innovation and hard work. And for insight into the 5G trends discussed above, read our latest annual 5G report for a deep dive on what’s happening on monetization and growth.
