It’s been an exceptionally busy time at Kaleido towers during Q1 2026 so far, with 3 major research efforts published that, when considered in tandem, formulate the basis for much of Kaleido’s continued research effort during the year.
Connectivity Growth High Among Regulated Markets
In January, we released our biannual Data Hub Update for H2 2025, which examines historical and future growth in cellular IoT connectivity across several domains, particularly vertical industries. More discussion on growth within specific IoT segments is available to Kaleido Members, including findings from our Quarterly Webinar. While much of the presentation is beyond the scope of this blog, it is important to provide some context, namely, the growth of connections in vertical industries:
Source: Kaleido Intelligence Connectivity Data Hub H2 2025
The first thing to note is that significant growth is expected across verticals such as transportation, energy and utilities, and healthcare. Notably, these are highly regulated industries, and how connectivity service providers address these verticals is becoming increasingly important: regulation, enterprise sophistication, and performance requirements linked to end customer business needs are eroding traditional connectivity models based on 901 IMSI ranges and limited infrastructure.
Enterprises Demand Policy Enforcement & Enhanced Orchestration
This hypothesis is validated through Kaleido’s research covering leading connectivity providers. Here, we conducted a 5-month deep dive into service provider capabilities across Connectivity Management, eSIM Connectivity, and eSIM subscription Management in our Connectivity Vendor Hub, which was published in February this year. The findings are quite telling, especially when we examine what service providers reported in terms of emerging vertical demands:
Source: Kaleido Intelligence Connectivity Vendor Hub 2025-2026
Notably, across several verticals, it is evident that complex orchestration, security, resilience and policy capabilities are required. In part, this is due to the fact that IoT is becoming integral to business operations. Proofs of concept have evolved into deployments that have scaled up, with a dual impact of forcing consideration of how service providers can address high availability and resilient scenarios.
Data Sovereignty & AI Require New Network Approaches
Perhaps the most significant issue is the regulatory environment, which now extends well beyond concerns related to permanent roaming and PII (Personally Identifiable Information). This forms much of the discussion in our latest report, Networks for IoT: Customer Demands and Service Provider Readiness. Here, we argue that data sovereignty requirements for enterprise IoT, as well as the growth of enterprise AI workloads, are transforming how network infrastructure must be designed. While AI adoption in production is still in the growth phase among enterprises, regulators around the globe are discussing how issues such as model training and inference must be handled, with the concept of ‘sovereign AI’ now a hot topic. This results in two key transformative requirements: on the one hand, service providers must be able to deploy infrastructure to meet diverse data sovereignty demands, as the environment today is highly fragmented. In the second instance, we discuss how the now-critical nature of IoT and the ability to support enterprise workloads depend on a resilient, programmatic network that legacy designs based on monolithic architectures will struggle to handle. This in itself provides a strong driver for 5G Standalone, which to date, has barely got off the ground in the IoT world.
The research provided a rather interesting opportunity to conduct a technical analysis of several players’ capabilities in terms of infrastructure and ability to address sovereignty, resilience and AI workload support. Cloud-native architectures provide obvious benefits, but key to tomorrow’s connectivity requirements will be how networks can support near-instantaneous failover, handle highly distributed termination or egress points for sovereignty, and bring the network for AI concepts away from the RAN and into the edge. This latter point is important: RAN-based AI suffers from many issues, including on-site power and cooling constraints, while inference points might not be ideally located. Metro edge can solve this, and provide a revenue opportunity for service providers – transitioning connectivity from merely a transport mechanism to an enterprise value asset.
The chart below highlights our findings, with positioning based on the ability to meet sovereignty and resilience requirements (x- and y-axis, respectively), and bubble sizes representing capabilities to support edge AI workloads.
Source: Kaleido Intelligence Networks for IoT Report
While the analysis does not provide a complete picture of the market, given the breadth of players in the space, it does attempt to bring together a collection of what Kaleido would classify as ‘innovators’ in the IoT connectivity ecosystem. There is deeper research to be done in this area, and we’ll no doubt dig into regulation, infrastructure and AI among other topics at MWC.
Blog Author: Steffen Sorrell (Chief of Research)