Research Finder
Find by Keyword
Can Hyperscalers Actually Run a Telco Network?
AWS's MWC 2026 blitz moves beyond simple cloud hosting to placing agentic AI and SaaS cores at the heart of carrier operations.
03/05/2026
Key Highlights
- Nokia and AWS launch agentic AI network slicing with du and Orange to automate real-time response to traffic surges.
- Ericsson debuts Agentic rApp as a Service on AWS Marketplace to push networks toward Level 4 autonomy.
- AT&T modernizes U.S. infrastructure by migrating core workloads to AWS Outposts and utilizing high-capacity fiber links.
- Liberty Latin America signs a five-year deal to deploy AWS Outposts for regional data sovereignty and IT modernization.
- Citymesh goes live with the world’s first commercial 5G Core SaaS, removing capital barriers for smaller operators.
The News
Amazon Web Services (AWS) has announced a series of strategic collaborations at Mobile World Congress 2026 that signal a shift from cloud-adjacent services to deep network integration. These partnerships with Nokia, Ericsson, AT&T, and Liberty Latin America focus on agentic AI for autonomous network management and the commercialization of 5G Core as a service. By deploying infrastructure like AWS Outposts directly into carrier facilities, AWS is positioning itself as the primary architectural layer for the next generation of connectivity. Find out more by clicking here to read the Nokia press release here, the Ericsson press release here, and the ATT press release here.
Analyst Take
We see a distinct change in the relationship between hyperscalers and telecommunications providers. For years, the conversation centered on whether telcos would be relegated to "dumb pipes" while cloud providers captured the value of the application layer. This year's announcements suggest a more intertwined reality where AWS is not just a vendor but the actual fabric of the network. The introduction of agentic AI is particularly telling. Unlike standard generative AI, which might just summarize a ticket, these agentic systems are architected to make decisions and execute changes in the radio access network without a human in the loop.
What was announced in these various collaborations represents a significant technical leap. The Nokia partnership introduces a 5G-Advanced network slicing solution that integrates Amazon Bedrock with Nokia’s MantaRay SMO and AirScale base stations. This system is designed to ingest real-world data, such as weather patterns and local event schedules, to automatically adjust Radio Access Network (RAN) policies. In a separate move, Ericsson launched its Agentic rApp as a Service on the AWS Marketplace. This tool uses a natural language interface designed to allow engineers to speak to the network to trigger complex optimization workflows. Technically, it connects to the Non-Real-time RAN Intelligent Controller via the Open RAN Alliance R1 interface, aiming to deliver what the industry calls Level 4 autonomy.
The AT&T collaboration is equally substantial in its physical scope. AT&T is migrating network service enablement and business support systems to AWS Outposts, which are essentially AWS-managed racks sitting in AT&T’s own data centers. This hybrid approach is designed to maintain the low latency required for network operations while gaining the flexibility of the cloud. Furthermore, AT&T is connecting AWS data centers with high-capacity fiber, creating a feedback loop where the carrier builds the highway that the cloud provider uses to manage the carrier.
We find the Citymesh deployment in Belgium to be the most disruptive of the bunch. It marks the first time a commercial mobile service is running on a 5G Core delivered entirely as Software-as-a-Service (SaaS). Architected to eliminate the traditional heavy capital expenditure of building a core, this model allows Citymesh to use a subscription-based approach to serve enterprises and SMEs. This is a classic "asset-light" strategy that could allow smaller players to compete with established giants by leveraging AWS’s global footprint for their core signaling and data management.
Liberty Latin America is taking a similar path but with a focus on regional sovereignty. Their five-year agreement involves deploying AWS Outposts in local facilities across the Caribbean and Latin America. This setup is designed to address the complex regulatory and data residency requirements of multiple island nations while modernizing over 500 customer workloads. We see this as a template for regional operators who lack the scale of a global tier-one carrier but still need to offer advanced AI-driven services to their enterprise clients.
The move toward agentic services is the clear thread here. We observe that AWS is no longer just providing virtual machines; it is providing the "brains" of the network. By integrating Amazon Bedrock with telecom-specific software from Nokia and Ericsson, AWS aims to deliver a level of automation that traditional legacy systems simply cannot match. This isn't just about efficiency; it's about creating new ways to charge for the network. If an operator can automatically slice its network to prioritize a stadium's payment systems during a surge, it can charge a premium for that guaranteed performance.
However, we must consider the risks of this deep integration. As carriers move their most sensitive core functions to the cloud, they become increasingly reliant on a single provider’s roadmap. While the use of Open RAN standards like the R1 interface is a positive sign for interoperability, the reality is that the orchestration layer is becoming an AWS-dominated environment. We are closely watching whether this leads to a new form of cloud lock-in that is even harder to escape than the hardware lock-in of the past decade.
From Infrastructure to Intellect: AWS and the Neural Overlay of Mobile Telecommunications
Moreover, from our perspective the AWS strategy revealed at MWC 2026 signifies a pivot from providing general-purpose cloud storage to becoming the neural overlay of global telecommunications, blending the line between software orchestration and physical layer operations. By embedding Amazon Bedrock directly into the RAN through Nokia’s MantaRay and Ericsson’s rApps, AWS is moving beyond basic automation into predictive infrastructure, where the network doesn't just react to congestion but anticipates it using non-telecom data such as local event schedules or weather patterns.
This Agentic-as-a-Service model represents a tectonic shift in the telco value chain: instead of carriers buying static hardware with decade-long lifecycles, they are subscribing to a dynamic, AI-managed network brain that resides on AWS Outposts within their own facilities. This can enable even mid-tier operators such as Liberty Latin America to bypass the massive R&D costs of autonomous networking, essentially renting Level 4 autonomy. Of key consideration, the technical dependency may prove unprecedented; as AWS manages the signaling, slicing, and optimization logic through the R1 interface, the carrier’s role risks transitioning from a technology owner to a local landlord for Amazon’s cognitive infrastructure.
Looking Ahead
Based on what we are observing, the telecommunications industry is currently navigating a fundamental transition from a connectivity-centric model to an orchestration-centric one. The key trend that we are going to be looking out for is the normalization of the "AI Factory" within the carrier environment. According to research from Deloitte, the global market for autonomous AI agents is projected to reach $8.5 billion by 2026, and the AWS announcements at MWC appear to be the first major industrial-scale manifestations of this projection in the telecom sector.
Going forward, we are going to be closely monitoring how the company performs on the delivery of Level 4 network autonomy. While the Agentic rApp approach from Ericsson aims to deliver a natural language interface for network management, the true test lies in the reliability of these agents under extreme network stress. When you look at the market as a whole, the announcement of a 5G Core SaaS with Citymesh suggests that the barrier to entry for new mobile virtual network operators (MVNOs) and private network providers has been permanently lowered.
HyperFRAME will be tracking how the company does in the competitive landscape against Microsoft Azure’s for Operators and Google Cloud’s Global Strategic Cloud Provider programs. While AWS has established a lead in agentic service orchestration, the battle for the "telco-native" cloud is far from over. Our perspective is that the successful carriers of the next decade will be those that can master the orchestration of these hyperscale tools without losing their fundamental identity as infrastructure owners. Would you like me to analyze the competitive responses from Azure or Google Cloud to these AWS announcements?
Ron Westfall | VP and Practice Leader for Infrastructure and Networking
Ron Westfall is a prominent analyst figure in technology and business transformation. Recognized as a Top 20 Analyst by AR Insights and a Tech Target contributor, his insights are featured in major media such as CNBC, Schwab Network, and NMG Media.
His expertise covers transformative fields such as Hybrid Cloud, AI Networking, Security Infrastructure, Edge Cloud Computing, Wireline/Wireless Connectivity, and 5G-IoT. Ron bridges the gap between C-suite strategic goals and the practical needs of end users and partners, driving technology ROI for leading organizations.
Share
Steven Dickens | CEO HyperFRAME Research
Regarded as a luminary at the intersection of technology and business transformation, Steven Dickens is the CEO and Principal Analyst at HyperFRAME Research.
Ranked consistently among the Top 10 Analysts by AR Insights and a contributor to Forbes, Steven's expert perspectives are sought after by tier one media outlets such as The Wall Street Journal and CNBC, and he is a regular on TV networks including the Schwab Network and Bloomberg.