Intelligent Infrastructure and the AI OpCo Moment: Takeaways from Mobile World Congress 2026

Mobile World Congress 2026 made one thing clear: Telecom is no longer debating whether AI will reshape the industry. It’s clear that it will. The questions now are who will build the intelligent infrastructure required to capture AI’s value, and who will operationalize it with telecom-grade trust.

This year, the dominant conversation is about intelligent infrastructure and what it takes to move from operating networks to operating intelligence. Across keynotes, roundtables, studio sessions and executive meetings, three themes kept converging: The industry still needs to finish the 5G and fiber journey; AI is becoming an operating layer rather than a feature; and trust — from fraud prevention to policy, sovereignty and security — is shifting to the center of telecom strategy. The next telecom wave is not primarily about moving more bits faster. It is about moving, governing and operationalizing intelligence.

For Snowflake, sponsoring the Intelligent Infrastructure track at MWC reflects where the industry is heading and why we are uniquely positioned to help telecom operators turn data into information that creates revenue beyond what growth alone can do. From keynote analysis to packed roundtables, studio sessions and executive meetings, Snowflake was front and center in the AI-native telecom conversation as a builder of the governed data foundations that autonomy requires.

Rather than treating AI as a feature layered on top of existing systems, intelligence becomes embedded into the core of the network itself. When data from networks, customers, operations and partner ecosystems can be unified, governed and then activated, it can power predictive optimization, better customer experiences and entirely new revenue streams. These include fraud prevention, location insights and programmable network capabilities. In short, telecom has an opportunity to evolve from being a connectivity provider to a platform that delivers trusted, operationalized intelligence at scale.

Why intelligence is becoming infrastructure

For decades, telecom operators have been in the business of moving data from Point A to Point B as reliably as possible. They built and maintained the lines that power everything from messaging and streaming to global commerce. But while the volume of data traveling those pipes continues to surge, revenue growth has not kept pace. That imbalance is forcing a strategic rethink. The conversation at Mobile World Congress 2026 made clear that the next chapter for telecom is about using that data to generate intelligence — on top of the continued buildout of stand-alone 5G, dense fiber, the enablement of nonterrestrial resilience, converged cloud-network architectures and increasingly distributed edge environments.

In a world of agentic AI, robotics, always-on inference and more distributed connectivity, intelligence behaves more like infrastructure. The shift from intelligence as application layer feature to intelligence as infrastructure was apparent at MWC across main-stage narratives and the conversations that matter most. In past years, AI was a matter of speculation and experimentation, with a question of whether to adopt it. This year showed that operators are asking what must happen for AI to run with telecom-grade reliability, and how to do that in ways that respect local regulation, existing platform investments and hybrid operating realities.

Intelligent infrastructure has requirements that feel familiar to anyone who has run a network at scale. It must:

• Execute under latency and determinism constraints

• Respect local regulation, data residency, security and policy

• Be anchored in identity, trust and governance

• Work across cloud, on-premises and edge environments without unnecessary data movement

• Integrate with the physical world, including devices, vehicles, sensors and industrial systems

That combination makes intelligence distribution look like a telecom problem, except the payload is not voice packets or internet traffic. It is decisions that realize tangible business outcomes.

This is the conceptual leap behind an AI OpCo: a network operator engineered to run intelligence as an operational workload, governed, metered and reliable, rather than as a feature bolted onto connectivity. It also reflects a broader shift visible across MWC: Telcos are being asked to become architects of a trusted environment, not just operators of transport.

From network operations to network intelligence

A consistent theme across executive conversations in Barcelona was the shift from reactive operations to predictive and increasingly autonomous systems.

If you strip away the buzzwords, intelligent infrastructure is a shift from network operations to network intelligence. It is the convergence of cloud, edge and network architectures into one coherent fabric capable of supporting AI at scale. It is the ability to orchestrate near real-time analytics and automation across hybrid environments so the network can move from reactive operations to predictive, and then to increasingly autonomous, decision-making.

In practice, that requires more than consolidating dashboards. It requires a unified, governed data foundation across network, service, customer, enterprise and ecosystem domains. It requires a semantic or knowledge layer that gives AI agents a consistent understanding of telecom context across vendors, systems and domains. And it requires an architecture that can span structured and unstructured data across multi-cloud and on-premises environments without creating a new generation of lock-in. One of the clearest messages from the week was that fragmentation — in data catalogs, tooling, ownership and access patterns — is now one of the main constraints on AI readiness.

These modern capabilities are broadly applied. The benefits show up in operational KPIs: lower incident resolution times, higher network resilience, improved service quality, better technician productivity and faster time to market. They also create new revenue pathways. Operators are exploring fraud and identity services, location verification, quality-on-demand guarantees, mobility insights, network APIs and industry-specific data products.

Telcos often underestimate the challenge of turning telecom data, documents and network capabilities into governed, consumable, billable products that developers and enterprises can adopt at scale.

That message came through clearly in the Intelligent Infrastructure programming at MWC. In our opening fireside chat, we focused on what must be true before networks can truly operate on “autopilot.” The conclusion was consistent with what many operators are learning the hard way: Autonomy does involve adding AI, but it also requires building the right data, governance and architectural foundations first. The sequence matters: Unify the data, establish trust and context, then automate. It also requires observability and validation, so operators can understand what agents are doing, test edge cases safely and scale with confidence.

Connecting intelligent infrastructure to business outcomes

At MWC, the conversation was pragmatic. AI is only strategic if it produces outcomes. Intelligent infrastructure is only useful if it produces measurable outcomes.

One of the clearest shifts this year was away from usage-only metrics and toward business and operational KPIs: time to insight, service quality, rollout velocity, recovered revenue, incident resolution, automated action rates and monetizable adoption. The three themes below dominated discussions with telecom leaders.

Create new revenue streams

Telecom operators sit on high-value signals: identity, location, device telemetry, quality metrics, network performance data and consented customer interactions. 

Monetization shifts from gigabytes to capability consumption: API calls, inference, data set subscriptions, quality-on-demand reservations, policy checks, private data sharing and clean-room collaboration. Revenue shifts from gigabytes to programmable value.

This also reframes how operators think about go to market. It is less about one-off integration fees and more about repeatable adoption and usage, with clear service levels, governance boundaries and distribution mechanisms that reduce friction for partners, developers and enterprises.

Improve operational efficiency

Telecom estates are complex by design: Legacy OSS and BSS systems, fragmented data domains, siloed operational processes and a mix of on-premises, multi-cloud and partner environments.

Unifying data across network, service, customer and enterprise domains reduces structural friction and enables modernization at scale. It accelerates decision-making, shortens the path from data generation to action, streamlines deployment patterns and reduces operating costs in environments where complexity has become the tax on every transformation initiative.

The qualitative impact is often as powerful as the quantitative gains. One Tier 1 operator team put it simply: “With Snowflake, you enable someone and there is silence after. Everything just works.” 

The most pragmatic path is often staged: Connect the highest-value structured data first, add unstructured context such as documents and field records, then layer agentic workflows on top. That is what makes AI-native data engineering so important. It turns modernization from a multiquarter integration exercise into a more repeatable and scalable operating model.

Operational simplicity matters because most telecom teams have learned to expect friction. Removing that friction is what makes step-function improvement possible.

Enable intelligent networks and operations

Proactive monitoring, predictive optimization and automated remediation require a unified, governed view of network truth.

As networks move toward higher levels of autonomy, the architectural ambition becomes clearer: Create a repeatable, reliable and deterministic ontology-based knowledge infrastructure for AI and agents. The network evolves from reactive infrastructure to an intelligent decision fabric. Telemetry becomes context. Context becomes decisions. Decisions become closed-loop action, with human oversight where needed.

This is where the ecosystem announcements at MWC matter. They are not side stories. They are signals of where the industry is building. Partners are helping operators add telecom-specific semantics, policy controls, observability, synthetic validation and governed agent frameworks so autonomy can scale without impacting auditability or explainability. That matters because the scenarios that matter most in telecom are often rare, sensitive or distributed across multiple domains. They have to be modeled, validated and governed before they can be automated with confidence.

EnterpriseWeb’s Snowflake Native App highlights how operators can implement autonomous networking agents with deterministic behavior and operational governance, extending Snowflake with a standards-based telecom ontology for explainable, logically consistent outcomes, auditability and policy controls. As Sreedhar Rao, Global Telecom CTO at Snowflake, put it: “EnterpriseWeb operationalizes telco data in Snowflake and provides a fast track to self-scaling, self-optimizing and self-healing networks.”

We also announced a joint solution with Rockfish Data to help operators and vendors validate autonomous network operations using privacy-preserving synthetic telemetry generated inside Snowflake’s governed environment. Validation is a critical gap in telecom: The scenarios that matter most are often rare, incomplete or too sensitive to share. Sreedhar captured the constraint and the opportunity: “Yet innovation has been constrained by limited access to realistic validation data. … Together with Rockfish, we are enabling carriers and vendors to generate high-fidelity test data within Snowflake’s governed environment — so they can move faster with confidence.”

Across partner momentum, on-stage discussions and our executive roundtable with leaders from Ericsson, Nokia and Cubic, the thread was consistent: The path to autonomy runs through governed data, semantic context, deterministic reasoning where needed and operational guardrails that make trust nonnegotiable.

Why Snowflake is in the right place at the right time

MWC is one of the few forums where global CSP executives, CTOs and heads of data strategy converge. The strategic conversations that happen there can compress months of executive alignment into days.

Snowflake’s presence and sponsorship of Intelligent Infrastructure signal our foundational participation in this transformation. Snowflake can be the control plane that connects network, cloud and ecosystem data, applies AI where the data lives, and makes it practical to productize trusted network signals as intelligent infrastructure — while protecting prior investments through interoperable, hybrid patterns. Snowflake helps telecom operators turn data into outcomes by:

• Creating a unified, governed data plane across network, cloud, enterprise and ecosystem domains

• Connecting structured and unstructured data through a shared semantic context that agents can understand

• Enabling privacy-first collaboration and monetization without unnecessary data movement

• Supporting AI inside a strong governance and security perimeter across hybrid and multi-cloud environments

• Providing distribution rails for platform-scale data products, applications, APIs and private shares

What stood out at MWC is that operators are not asking for another disconnected AI layer. They are asking for a governed, interoperable foundation that works with their existing cloud commitments, their on-premises realities, their data residency obligations and their need to prove business value quickly.

A platform business may rely on bespoke integrations, but it scales through repeatable patterns, clear contracts and distribution mechanisms that reduce friction for partners and customers.

The bigger shift: From utility to platform

Perhaps the most significant takeaway from MWC 2026 is that telecom’s future will be defined more and more by programmable capability.

In the AI era, telcos can remain as connectivity pipes optimized for cost per bit, or become AI OpCos running intelligent infrastructure, where connectivity becomes capability and the network becomes a trusted fabric for data-driven decisions. That future is also more software-driven, more standards-aware and more ecosystem-led. It depends on operators exposing capabilities through APIs, data products and governed sharing models rather than treating every monetization path as a one-off project.

Growth will increasingly be measured in adoption, API calls, inference consumption, quality guarantees, data sharing and ecosystem integration. The operators who win will look more like platform companies than utilities. They will monetize network truth as products.

The road ahead

The industry is at a structural inflection point. Intelligent infrastructure is the lever that can deliver platform-like returns, where traffic growth alone can’t.

Snowflake is not the entire telecom stack. Operators will still need edge compute, orchestration and network modernization. They will still need to continue the 5G journey, strengthen the fiber, enable nonterrestrial resilience, and cloud foundations beneath it, and manage a more distributed future of connectivity. They will also need strong governance around security, fraud, regulation and policy as AI reaches deeper into operations. But without a unified, governed data and AI foundation, those investments cannot fully translate into monetizable intelligence.

It’s time for telecom leaders leaving MWC 2026 to think less about whether AI will matter, and more about whether they are building the intelligent infrastructure, knowledge layer and operating model required to capture the value it creates. The ones that do will move beyond connectivity and toward becoming AI OpCos delivering intelligence reliably, locally and with telecom-grade trust.