Our latest news and insights
The latest news and insights
Why Sovereign AI Is Forcing Operators to Rethink the Network Core
Sovereign AI in Telecom: Control, Governance, and the Race for Architectural Independence
A new TM Forum report just put numbers to something operators have been sensing for a while: the race to adopt AI in telecom isn't just about technology. It's about control.
According to TM Forum research published ahead of DTW Ignite 2026, 72% of communication service providers (CSPs) are already investing in or exploring sovereign AI. Nearly all — 98% — are adapting or exploring changes to their network architecture. And 96% are aligning more closely with national or regional policy frameworks.
These aren't marginal numbers. They signal that the telecom industry has moved past the question of whether AI belongs in the network. The question now is who governs it — and what that means for the core.
What Is Sovereign AI in Telecom?
Sovereign AI refers to a country's or organization's ability to develop, deploy, and govern AI systems using infrastructure and data they control — without strategic dependence on foreign vendors or platforms.
For telecom operators, the concept lands directly in the network core. TM Forum President Nik Willetts and Chairman Steffen Roehn frame it clearly in their discussion paper Making Sovereign AI Real:
"Sovereign AI will not be made real by owning GPUs or building data centers. It will be made real by controlling the enforcement layer that governs how AI behaves in motion — across every interaction, in real time. Regulators around the world are moving fast to make this a legal requirement."
That enforcement layer runs through the core network. Which means architectural decisions made today — about modularity, vendor dependencies, and cloud-native readiness — will determine how much sovereignty operators actually have tomorrow.
The Regulatory Context Is Accelerating This Shift
The regulatory pressure reinforcing sovereign AI in telecom is real and growing:
- The EU AI Act requires telecom and technology companies to manage AI according to regional laws, with traceability and compliance obligations tied to where and how AI operates.
- The EU's proposed Cloud and AI Development Act (CADA) goes further, positioning sovereign AI as an auditable legal requirement — not just a strategic preference.
- In the United States, a patchwork of executive orders and sector-specific enforcement is pushing in the same direction. Executive Order 14409, Promoting Advanced Artificial Intelligence Innovation and Security, introduces pre-release review frameworks for advanced AI models shared with federal agencies.
The direction is consistent across geographies: operators will increasingly need to demonstrate that their networks support localized AI governance, data residency, and control over how intelligent systems behave at the core level.
The Vendor Lock-In Problem Gets More Expensive
Analyst Vish Nandlall, writing for RCR Wireless, identified a critical bottleneck in how many countries are approaching sovereign AI: they are subsidizing infrastructure they don't control, and purchasing access to "someone else's silicon, someone else's models, and someone else's operating layer."
The same risk applies at the operator level. A core network built around a single vendor's proprietary stack may appear stable — until that vendor's roadmap diverges from yours, until integration becomes prohibitively expensive, or until regulatory requirements demand architectural changes the vendor isn't prepared to support on your timeline.
Vendor lock-in in telecom is not a new problem. But in the context of sovereign AI, the cost of it has changed. When the core network becomes the enforcement layer for AI governance, losing architectural control isn't just a commercial inconvenience — it's a strategic exposure.
Why Cloud-Native, Modular Architecture Is the Answer
Modular, open, cloud-native core network architectures don't just reduce vendor lock-in risk. They make it manageable on the operator's schedule — not the vendor's.
Concretely, this means:
- Independent function evolution. A modular core allows operators to upgrade or replace individual functions — SDM, Policy, IMS, SMSC — without touching the rest of the stack. New capabilities can be introduced without a full rip-and-replace cycle.
- Progressive AI adoption. Cloud-native architectures enable operators to integrate AI-driven automation incrementally, starting where the business case is clearest, and expanding without being locked into a single AI platform or vendor ecosystem.
- Regulatory readiness. A modular, standards-based core makes it significantly easier to demonstrate data residency, localized control, and AI governance compliance to regulators — because the architecture is transparent and auditable by design.
- Multivendor interoperability. True multivendor strategies are only possible when the core is built on open interfaces. Operators who have avoided proprietary dependencies are far better positioned to select best-of-breed components as the AI-native network evolves.
What This Means for Operators Evaluating Their Core Strategy
For operators considering how to position their network for sovereign AI demands, a few strategic questions are worth pressure-testing:
- Can you add or replace individual core functions without touching the rest of the stack?
- Can you deploy AI-driven automation incrementally?
- Can you demonstrate to regulators that your network supports localized AI governance and data sovereignty?
- Are you on your vendor's roadmap — or are they on yours?
If the answer to any of these is uncomfortable, the architecture — not the AI — is the problem.
Summa Networks and the Modular Core Approach
At Summa Networks, we have spent over a decade helping mobile operators, IoT networks, and private network operators across more than 30 countries build core infrastructure that evolves on their terms.
Our Full Core portfolio — including Unified SDM, Extended Converged Control Plane, IMS Core, Policy Function, and SMSC — is designed from the ground up for modularity and interoperability. Operators can adopt individual components, integrate with existing infrastructure, and expand capabilities without changing providers or committing to a single vendor ecosystem.
In a world where 98% of CSPs are rethinking their network architecture, that kind of architectural freedom isn't a differentiator. It's a prerequisite.
Key Takeaways for Sovereign AI in Telecom
- Sovereign AI in telecom is not about owning hardware — it's about controlling the enforcement layer in the network core.
- 72% of CSPs are already investing in sovereign AI; 98% are rethinking their network architecture (TM Forum, 2026).
- Regulatory frameworks (EU AI Act, CADA, US EO 14409) are making sovereign AI governance an auditable requirement.
- Vendor lock-in in a sovereign AI context is a strategic and regulatory risk, not just a commercial one.
- Cloud-native, modular core architectures are the structural answer — enabling progressive AI adoption, regulatory compliance, and genuine multivendor flexibility.
Sources:
- TM Forum – "Making Sovereign AI Real" (2026): newsroom@tmforum.org
- RCR Wireless AI Infrastructure Daily, June 19, 2026:
https://content.rcrwireless.com/june-19-2026-ai-infrastructure-daily - EU AI Act:
https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence
Latest Blog
Cloud-native SDM for 4G and 5G Networks with 2G Integration
