As AI adoption accelerates, telecom operators have a chance to claim a central role in the AI economy, says Spirent (now part of Keysight). To seize it, they must move beyond connectivity and deliver trusted, high-performance AI infrastructure and services.
The AI boom is real—and accelerating. In a recent Bain & Company survey, 95% of U.S. companies said they are now using generative AI. And in its 2025 AI in the Workplace report, McKinsey forecasts that 92% of companies plan to increase their AI investments over the next three years.
For service providers, this is more than just the latest tech craze. It’s a once-in-a-generation opportunity to redefine telecom’s role in the digital economy and claim a central position in the AI value chain.
So, how can operators seize this moment and lead? As we work with service providers developing new AI service offerings worldwide, here’s what we’re hearing.
Telco’s big AI bet
Telcos have not been idle in the AI race; they’ve been actively investing. Much of this effort has focusing on applying AI to their own internal operations—from fraud detection to network automation to predictive maintenance. But the larger opportunity lies on the customer-facing side, where Analysys Mason estimates operators are investing more than $21.6 billion over the next three years to develop new AI service offerings, especially for governments and enterprises.
We see four main areas where telcos are positioning to monetize AI:
- AI networking as a service: With their unpredictable traffic patterns and extreme performance demands, AI workloads are not like traditional internet traffic. Telcos can offer premium connectivity services tailored to meet these requirements under service-level agreements (SLAs).
- AI infrastructure as a service: Telcos already own data centers and distributed edge infrastructure and have longstanding relationships with public cloud providers and regional colocation data centers. They’re well positioned to repurpose their infrastructure resources as AI-optimized hosting environments—especially for sovereign or private AI models that require strict data governance.
- AI models as a service: Enterprises increasingly need help navigating the vast and growing number of large language models (LLMs) and other AI models. Telcos can act as brokers or orchestrators, connecting customers to the right domain-specific models while handling compliance, security, and integration on their behalf.
- AI applications: As AI applications strain traditional transport networks, telcos are among the first to optimize their infrastructures for AI requirements. Now, some are commercializing their AI-optimized networking capabilities as industry-specific solutions.
The sovereign AI inflection point
Within those models, operators are focusing on consumers, enterprises, and governments. For consumers, the opportunity is largely in aggregation, with some operators exploring curated, app-store-style platforms.
However, most operators see enterprise and government customers as the larger, more exciting opportunity.
Many enterprises, especially those in regulated industries, want the benefits of AI without ceding control of their data or relying on opaque public cloud models. Telcos can offer the infrastructure, compliance, and performance guarantees these customers require—especially in industries like finance, healthcare, and manufacturing.
Similarly, as new regulations governing data use, model training, and algorithmic transparency emerge, many governments are unable to use multinational hyperscale data centers for AI workloads. They need infrastructure that meets regulatory requirements, ensures data sovereignty, and delivers greater local control, security, and auditability. Once again, telecom is uniquely positioned to meet these needs.
Telcos already have the physical infrastructure to meet sovereign AI demands—low-latency networks, edge compute, secure transport—and the regulatory know-how to operate within national frameworks. Many already operate data centers that can be adapted to host sovereign models. But to be credible in this space, telcos must demonstrate that their networks can meet stringent AI networking and compliance requirements, even as they continue to evolve.
Testing matters more than ever
Telcos must meet a long list of requirements to fully capitalize on the AI opportunity. They need to ensure and demonstrate to customers that AI offerings:
- Meet stringent data sovereignty and governance requirements: Sovereign AI services must be validated to comply with relevant regulations such as the National Institute of Standards and Technology Cybersecurity Framework (NIST CSF) or General Data Protection Regulation (GDPR). To build trust, AI models should be continuously audited for accuracy, inference efficacy, and transparency.
- Protect sensitive data and infrastructure: Government and enterprise applications demand stringent security and reliability. Operators must validate that the right security frameworks, systems, and policies are in place, and that the infrastructure is operationally resilient.
- Meet variable and extreme performance requirements: Different AI applications like image and video generation, computer vision, recommendation engines, and others bring very different transaction sizes and traffic patterns. AI workloads can also be highly dynamic and require extremely low-latency, lossless performance. Operators must continually validate that AI offerings meet these requirements and can scale without compromising data residency rules or SLAs.
- Minimize business and legal risk: Todeliver custom AI services under SLAs, telcos must have comprehensive test and assurance processes in place. That includes proactive testing, so they can catch performance and security issues early and avoid penalties for themselves or their customers.
Fortunately, many telcos have already taken significant steps to address these requirements, embedding test and assurance into every phase of service delivery. Key elements of these efforts include:
- Digital twins, which enable operators to simulate failure scenarios, traffic surges, and AI workload stress in realistic virtual environments, without risking production traffic
- Active assurance, where operators inject synthetic test traffic into live networks to measure real-time performance, uncover anomalies, and continually validate service-level compliance
- Continuous testing, with more operators integrating automated testing capabilities into continuous integration/continuous delivery (CI/CD) pipelines so they can continually validate that AI services are performing as expected
Looking ahead
AI represents more than a new technology cycle. It’s a fundamental shift in how value is created and delivered. For service providers, it’s also a rare opportunity to reposition telecom as a central enabler in the AI value chain.
But this opportunity won’t be realized without execution. By combining their unique strengths and resources with continuous testing and assurance, telcos can transform from bandwidth provider to strategic AI enabler.
