Palo Alto Networks: Securing AI Data Centres by Design

Palo Alto Networks has unveiled an expanded security ecosystem aimed at protecting what it calls ‘AI Factories’, positioning them as the operational core of the emerging AI economy.

As organisations accelerate investment in large-scale model training and inference environments, the company argues that security must be engineered into AI infrastructure from the outset rather than layered on later.

Announced at Mobile World Congress 2026 in Barcelona, the initiative brings together collaborations with Nokia, U Mobile, Aeris and Celerway. The programme is designed to secure both the physical and digital layers underpinning AI, from hyperscale data centres to telecom networks and distributed edge sites that feed data into AI platforms.

As AI workloads drive demand for multi-terabit throughput and densely packed compute clusters, operators are reassessing more than power and cooling. Embedded security across networks, workloads and connected devices is becoming central to AI deployment strategies.

“We are establishing the secure foundation for the AI economy through extensive ecosystem collaboration,” says Anand Oswal, Executive Vice President at Palo Alto Networks.

“By seamlessly integrating our AI-powered security services directly from the data centre into the most vital 5G and IoT networks globally, we are ensuring the AI Factory is secure by design.

“These partnerships enable us to create a secure digital infrastructure capable of managing the multi-terabit throughput required for training AI models.”

Sovereign and secure AI data centres

A key strand of the announcement focuses on sovereign AI and the development of large-scale European AI facilities.

Through its partnership with Nokia, Palo Alto Networks is combining security platforms with AI-optimised data centre infrastructure to support the rise of AI gigafactories across Europe. The joint architecture is intended to address data sovereignty requirements while maintaining performance and scalability.

As European operators invest in AI capacity, extending security controls from the network layer into applications and AI workloads is becoming a strategic priority.

“In the race to build the world's AI Factories, you cannot leave the door open at the infrastructure layer,” says Greg Dorai, Senior Vice President and General Manager, IP Networks at Nokia.

“Nokia and Palo Alto Networks jointly envision comprehensive architectural and operational frameworks that expand security solutions from the network layer to workloads.

“The validated architecture will allow our customers to build future-proof, sovereign data centres. We aren't just providing connectivity, we are protecting the physical and digital integrity of industrial digitisation at scale.”

Extending AI security from core to edge

Beyond centralised facilities, the ecosystem extends into 5G and IoT environments that generate and transport data used by AI systems.

New agreements with U Mobile, Aeris and Celerway Communication broaden the scope from core infrastructure to connected devices and mobile edge deployments.

In Malaysia, Palo Alto Networks has signed a memorandum of understanding with 5G provider U Mobile to collaborate on a network-embedded Security as a Service model.

By integrating next-generation firewalls and AI-driven protection directly into 4G and 5G networks, the partners aim to provide built-in security for both enterprises and consumers.

Through its partnership with Aeris, the company is addressing large-scale IoT environments. Integration between Aeris IoT Watchtower and Prisma SASE 5G enables organisations to apply zero trust policies and data loss prevention controls across millions of wireless devices from a unified control point.

For sectors such as healthcare, manufacturing, retail and utilities, this helps reduce the expanded attack surface created by connected devices supplying data into AI pipelines.

Celerway Communication strengthens protection at the distributed edge. By integrating Palo Alto Networks VM Series Next Generation Firewalls into 5G edge deployments, first responders and remote teams can maintain encrypted data integrity and a consistent security posture, even in highly mobile or remote settings.

Embedding security into AI architecture

Taken together, the ecosystem reflects a broader shift in how AI infrastructure is being designed. High-performance clusters, liquid cooling and dense interconnects remain essential, but securing training pipelines, inference services and data flows from edge to core is increasingly seen as foundational.

By aligning data centre platforms with telecom networks and IoT management layers, Palo Alto Networks and its partners are seeking to embed security directly into AI builds.

For organisations pursuing sovereign AI strategies, the emphasis is on validated architectures that unify network, workload and edge protection within a single, integrated framework.