The Trust and Intelligence Layer: How Agentic AI and Sovereign Infrastructure Are Redefining Telecom

The telecom industry is moving toward an AI-native operating model in which agentic AI and sovereign infrastructure reinforce one another. As operators deploy AI agents across network operations, service orchestration, operations support systems/business support systems environments, and customer-facing processes, trusted compute, local data governance, resilient architectures, and quantum-safe security become operating requirements rather than supporting capabilities.

Avasant’s Telecom Digital Services 2026 Market Insights™ identifies three converging themes: agentic AI and autonomous networks, sovereign AI infrastructure, and resilient digital trust. Their convergence matters because each new layer of AI infrastructure creates additional requirements for operational automation, governance, and security.

At the same time, governments and enterprises are accelerating investments in AI factories, domestic data centers, sovereign cloud environments, and trusted AI ecosystems. Telecom operators can support this transition as infrastructure providers and operators of trusted digital environments, as they own national networks, edge infrastructure, data center assets, and enterprise connectivity. However, becoming providers of both the trust layer and the intelligence layer will depend on their ability to govern autonomous decisions, operate competitive compute services, and protect data and models across distributed environments.

Agentic AI Is Advancing Networks from Predefined Automation to Governed, Autonomous Execution

One of the most significant developments reshaping telecom operations is the shift from predefined automation, which follows fixed rules and workflows, to agentic systems that analyze changing conditions, select actions, and initiate execution across network domains. Operators are increasingly deploying AI agents to predict faults, optimize resources, identify service impacts, and initiate corrective actions across RAN, transport, and core environments. The defining change is the transfer of bounded decision-making authority from predefined workflows to governed AI agents.

Available industry surveys indicate growing investment, but they do not yet demonstrate broad operational autonomy. TM Forum reports that the share of operators reaching Level 3 or higher autonomous network maturity increased from 19% in 2025 to 21% in 2026, while approximately 75% of communication service providers plan to increase investment[1]. NVIDIA’s telecom AI survey separately reports that 65% of operators use AI-driven network automation and that 50% identify autonomous networks as the highest-return AI use case[2]. Because these measures use different samples and definitions, they should be read as directional indicators rather than a combined measure of market maturity.

As telecom environments become increasingly distributed across AI-enabled RAN architectures, edge infrastructure, cloud environments, and sovereign AI ecosystems, operators require greater levels of automation, orchestration, and real-time decision making. Agentic AI is emerging as an operational layer that enables networks to analyze conditions, optimize resources, coordinate actions, and respond to issues with minimal human intervention.

This shift is also visible across the service-provider landscape. Based on submissions reviewed for Avasant’s Telecom Digital Services 2026 RadarView™, around 57% of participating providers reported dedicated platforms that combine AI, analytics, and automation to support closed-loop network operations and service assurance. This finding reflects the participating provider sample rather than the wider market, but it indicates that providers are productizing capabilities in anticipation of operator demand for automated decision-making and increasingly autonomous network operations.

The examples below combine announced initiatives and reported operational outcomes. They should be assessed by deployment status, production scale, degree of human oversight, and whether the stated benefits have been independently verified. Each quantitative claim should, therefore, be linked to a dated primary source.

Several operators have already begun deploying these capabilities at scale[3]:

Taken together, these cases show that operators are applying autonomous approaches to selected workflows spanning monitoring, assurance, incident response, and workload recovery. They do not yet establish consistent, end-to-end autonomy across network domains; the stronger evidence concerns bounded processes that can be monitored and escalated to human operators.

Telecom Operators Are Becoming Key Enablers of Sovereign AI Ecosystems

Governments are increasingly treating AI compute as strategic infrastructure as AI has become critical to economic competitiveness, national security, critical public services, and digital sovereignty. Growing concerns about data control, access to compute capacity, and reliance on external AI ecosystems are driving investments in domestic AI infrastructure.

Public policy is strengthening the case for domestic AI capacity. The European Union’s InvestAI initiative intends to invest €200 billion in AI[4], while Canada’s Sovereign AI Compute Strategy supports domestic infrastructure, compute capacity, and sovereign AI development[5]. These programs signal government intent, but their impact on telecom operators will depend on the funding mix, allocation mechanisms, implementation timelines, and the extent of operator participation.

As governments and enterprises seek greater control over AI workloads, data governance, and digital infrastructure, telecom operators can support sovereign AI ecosystems through national networks, edge infrastructure, interconnection assets, data center facilities, and established enterprise relationships. These assets create a potential advantage in regulated and latency-sensitive workloads, but they do not guarantee competitiveness. Operators must also secure access to power and accelerators, build cloud and model partnerships, strengthen platform engineering and compliance capabilities, and develop commercial models that can compete with hyperscalers and specialist data center providers.

The initiatives below vary substantially in status and scope. Some represent committed infrastructure, while others describe partnerships, capacity targets, or strategic programs. Comparison, therefore, requires consistent disclosure of announcement date, deployment stage, capital committed, infrastructure ownership, and the share of capacity reserved for sovereign workloads.

Several operators have already begun making significant investments[6]:

These investments expand the role of selected operators in domestic compute, secure connectivity, and trusted hosting. Their strategic value will depend on whether operators can translate infrastructure announcements into differentiated services, sustained utilization, regulatory trust, and acceptable returns on capital.

The Feedback Loop: Sovereign AI Infrastructure and Agentic AI Reinforce One Another

Sovereign AI infrastructure and agentic AI reinforce one another: distributed AI infrastructure increases operational complexity, while agentic systems provide the orchestration needed to manage that complexity.

As governments, enterprises, and operators invest in sovereign AI factories, GPU clusters, AI-ready data centers, distributed edge environments, and AI-enabled network architectures, telecom environments are becoming larger, more distributed, and increasingly interconnected. Such ecosystems require continuous coordination of workloads, network resources, data flows, service assurance processes, and security controls across multiple infrastructure layers.

As AI workloads spread across AI factories, cloud environments, enterprise platforms, and edge locations, operators must coordinate traffic, compute, data movement, service assurance, and security across a larger distributed footprint. The International Telecommunication Union’s ION-2030 framework identifies AI-enabled networks, distributed intelligence, and autonomous operations as building blocks of next-generation digital infrastructure[7].

The growing complexity of these interconnected environments is putting pressure on traditional operating models, driving operators toward agentic AI, autonomous orchestration, self-healing networks, and closed-loop automation.

The resulting feedback loop is both operational and strategic. Sovereign cloud environments, AI factories, edge deployments, and distributed compute ecosystems create more infrastructure and workload dependencies for operators to coordinate. In turn, agentic orchestration, closed-loop automation, and governed autonomous decision-making become necessary to operate these environments efficiently at scale.

Trust Requires Resilience: Quantum-Safe Security Is Emerging as a Strategic Priority

As governments and enterprises invest in sovereign AI ecosystems, the focus is expanding beyond domestic compute capacity and data residency to include long-term security and digital trust. Sensitive enterprise information, government data, and AI models hosted within sovereign environments must remain secure throughout their life cycle, making resilience an increasingly important element of sovereign AI strategies.

This shift is reflected in growing efforts to prepare for the quantum era. US federal guidance requires agencies to inventory cryptographic systems and migrate toward NIST post-quantum cryptography standards, while CISA and other agencies are supporting critical-infrastructure readiness[8]. The European Union’s post-quantum road map similarly calls for critical infrastructure to transition by 2030[9]. Together, these measures respond to the risk of “harvest now, decrypt later” attacks, in which encrypted data collected today could be decrypted as quantum computing capabilities mature.

For telecom operators, these developments are closely linked to the increasing role of networks in supporting AI factories, government workloads, enterprise AI platforms, and distributed edge environments. As these ecosystems continue to expand, the need to maintain resilient networks, protect data exchanges, and support trusted digital communications will only increase. As a result, quantum-safe security and resilient network architectures are becoming important factors in the design of future telecom infrastructure.

Operators have begun incorporating post-quantum cryptography, quantum-safe networking, trusted hosting, and sovereign control into their infrastructure strategies. These capabilities are related but not equivalent: post-quantum controls address cryptographic migration, while sovereign hosting and resilient network design address jurisdiction, operational control, and service continuity.

Several operators have already begun deploying these capabilities at scale[10]:

As sovereign AI ecosystems continue to expand, trust will increasingly depend not only on where AI workloads run, but also on how securely data, models, and digital infrastructure are protected across their life cycle.

The Way Forward

Telecom leaders should treat sovereign AI infrastructure and agentic AI as a coordinated portfolio rather than as separate investment programs. The portfolio should be governed against three outcomes: trusted infrastructure capacity, bounded autonomy in network operations, and resilient security across data, models, and connectivity.

Operators should first identify the workloads for which sovereignty, latency, resilience, or regulatory control creates a defensible advantage. They can then align sovereign cloud environments, domestic data centers, edge platforms, and secure interconnection with those workloads, while expanding agentic orchestration, closed-loop automation, and self-healing capabilities in network domains where governance and escalation controls are mature.

Security and resilience should be designed into infrastructure programs from the outset. Operators should inventory cryptographic dependencies, prioritize long-lived and sensitive data for post-quantum migration, define governance for AI agents and models, and test recovery across sovereign cloud, edge, and network environments. These controls are particularly important where operators support government and regulated enterprise workloads.

Looking ahead, sovereignty is likely to extend beyond compute, data, and infrastructure to encompass trust itself. As AI becomes embedded in critical sectors and national digital ecosystems, the ability to provide secure, resilient, and quantum-ready infrastructure may become as important as providing AI compute capacity. Telecom operators are uniquely positioned at this intersection, connecting AI factories, cloud environments, enterprises, governments, and edge locations through networks that increasingly underpin digital economies.

The opportunity is not simply to add AI compute to a connectivity portfolio. Telecom operators can differentiate by governing how sovereign AI workloads are connected, operated, and protected across national, cloud, enterprise, and edge environments. Those that combine credible infrastructure economics with bounded agentic autonomy and demonstrable resilience will be better positioned to provide the trust and intelligence layers of AI-native economies.

By Jatin Gulati, Research Analyst, and Sahaj Kumar, Research Director, Avasant

References

[1] https://inform.tmforum.org/research-and-analysis/reports/a-regional-guide-to-autonomous-networks-progress

[2] https://www.nvidia.com/en-us/lp/industries/telecommunications/state-of-ai-in-telecom-survey-report/

[3] https://www.telekom.com/en/media/media-information/archive/ai-agents-for-mobile-network-1099054
https://inform.tmforum.org/research-and-analysis/case-studies/china-mobile-achieves-level-4-an-in-network-operation-center-with-intelligent-agents
https://www.telefonica.de/news/press-releases-telefonica-germany/2026/03/o2-telefonica-develops-ai-assistant-for-network-operations-noa-is-a-digital-sparring-partner-for-network-engineers.html
https://itwire.com/it-industry-news/strategy/telstra-advanced-autonomous-networks-ambition-through-break-through-collaboration-with-red-hat,-dell-technologies-and-cisco

[4] https://ec.europa.eu/commission/presscorner/api/files/document/print/en/ip_25_467/IP_25_467_EN.pdf

[5] https://stip.oecd.org/stip/interactive-dashboards/policy-initiatives/2025%2Fdata%2FpolicyInitiatives%2F99998538

[6] https://www.airtel.in/press-release/10-2025/airtel-partners-with-google-to-establish-indias-first-mega-ai-hub-and-data-center-in-visakhapatnam/
Airtel Announces US$1 Billion Investment in Nxtra Led by Alpha Wave Global and Existing Investor Carlyle. Bharti Airtel Will Also Participate
https://www.telekom.com/en/newsroom/latest-updates/media-information/2025/11/for-a-sovereign-germany
https://news.sktelecom.com/en/2056
https://news.sktelecom.com/en/3124
https://aiforgood.itu.int/how-china-telecom-is-building-intelligent-infrastructure-for-ai/
https://www.ctyun.cn/products/ctxirang
https://www.rcrwireless.com/20250808/ai-infrastructure/orange-ai
https://www.eand.ae/en/enterprise-and-government/ai-solutions/ai-net.html

[7] https://prodnew.eandenterprise.com/en/press-release/e-enterprise-launches-sovereign-inference-ai-platform-in-country-ai.htmlhttps://www.itu.int/hub/2026/02/ion-2030-enabling-the-intelligent-optical-network-era/

[8] https://www.whitehouse.gov/presidential-actions/2026/06/securing-the-nation-against-advanced-cryptographic-attacks/

[9] https://digital-strategy.ec.europa.eu/en/library/coordinated-implementation-roadmap-transition-post-quantum-cryptography

[10] https://www.orange-business.com/be-en/solutions/cloud/sovereign-cloud
https://news.sktelecom.com/en/2056
https://www.singtel.com/business/products-services/quantumsafenetwork
https://prodnew.eandenterprise.com/en/press-release/e-enterprise-launches-sovereign-inference-ai-platform-in-country-ai.html
https://www.telekom.com/en/newsroom/latest-updates/media-information/2025/11/for-a-sovereign-germany