The race for Artificial Intelligence is forcing companies to make a decision that goes beyond performance: where the data resides and under which jurisdiction it is processed. In this context, Hewlett Packard Enterprise (HPE) and 2degrees, one of New Zealand’s leading telecommunications and technology providers, have announced a strategic collaboration to deploy a private AI platform with an explicit goal: to accelerate innovation without sacrificing control, security, and data sovereignty.
The agreement, announced from Auckland on January 26, 2026, is based on HPE Private Cloud AI, a “turnkey” solution developed by HPE in partnership with NVIDIA within the NVIDIA AI Computing by HPE initiative. In practice, the plan aims to modernize 2degrees’ tech environment with a private, unified architecture capable of scaling compute and storage resources as demand evolves, while maintaining the premise that critical customer and operational data stay within New Zealand territory.
AI enters the network, but with data “at home”
This announcement responds to an increasingly common tension in regulated or critical infrastructure sectors: AI promises efficiency and automation, but also multiplies the volume of sensitive data moving through systems, clouds, and supply chains. For a telecom, where network data and customer data are core to the business, the dilemma is even more intense.
In its statement, HPE emphasizes that the new platform will allow 2degrees to dynamically allocate resources among various AI use cases while maintaining compliance with sovereignty requirements. The company’s goal is to improve reliability, reduce outages, and optimize network performance through an integrated on-premise environment that combines secure data management and advanced AI-driven analytics.
This emphasis on “onshore” isn’t accidental. In New Zealand, as in other markets, the conversation about data sovereignty has intensified as cloud services and data exchanges across multiple jurisdictions grow. Government guidance on jurisdictional risk in the cloud and privacy frameworks reinforce the idea that data control isn’t just a technical issue, but also a legal and strategic one.
An “AI factory” ready to operate
HPE defines the proposal as an AI factory architecture: a preintegrated block designed to reduce friction in enterprise deployments, accelerate development cycles, and facilitate teams to move from isolated pilots to production use cases.
For 2degrees, the focus is on network operation and moving towards more automated models. The company states that the combination of HPE infrastructure and software, along with NVIDIA Enterprise support and AI technologies, will enable more resilient, predictable, and scalable networks for their customers. At the same time, the platform aims to speed up the rollout of new products and features, with greater responsiveness to market changes.
Colin Henderson, CEO of HPE New Zealand, frames the initiative as an example of “responsible” adoption: a way to drive AI transformation without losing control of the data. In his words, deploying this architecture at 2degrees “sets a benchmark” for how New Zealand organizations can incorporate AI while maintaining local data governance.
Use cases: autonomous network, predictive maintenance, and capacity planning
Beyond concepts, the collaboration comes with a concrete list of initial applications. HPE and 2degrees point to three areas with direct impact on costs and customer experience:
- Autonomous network operations, automating tasks and supporting operational decision-making.
- Predictive maintenance, anticipating failures and degradations before they escalate into incidents.
- AI-driven capacity planning, to adjust investment and resources based on real demand patterns.
Stephen Kurzeja, CTO of 2degrees, states that HPE and NVIDIA’s private AI platform will allow “faster progress,” extracting greater value from data and building solutions that make the network smarter and more robust. Kurzeja also highlights a key shift for 2026: the adoption of multi-agent use cases, scenarios where multiple AI agents work in coordinated tasks for operation and network optimization, moving toward a more “self-driving” infrastructure.
The focus on local agents has technical and sovereignty implications: running inference and analytics close to the data, with access control and without relying on continuous transfers to external clouds, can improve latency, reduce attack surfaces, and simplify audits.
Data sovereignty as a competitive edge
Although the announcement is framed as a technological advance, its business interpretation is clear: data sovereignty is becoming a competitive argument. For a telecom, proving that customer and operational data is stored and governed under local jurisdiction can reduce regulatory uncertainty, ease agreements with sensitive agencies and partners, and strengthen trust at a time when AI is amplifying concerns over privacy and cybersecurity.
At its core, this also signals to the market: AI isn’t just moving toward the public cloud. Simultaneously, demand is increasing for private platforms and on-premise deployments that enable organizations to capture value without ceding control. In sectors like telecommunications, banking, government, or healthcare, balancing performance, cost, compliance, and sovereignty is beginning to shape the roadmap.
Frequently Asked Questions
What is a private AI platform with data sovereignty, and how does it benefit a telecom?
It’s an AI infrastructure deployed on private (on-premise) environments that keeps data within local jurisdiction. For a telecom, it’s used to analyze and optimize network operations, automate processes, and improve customer experience without exposing critical data to unnecessary transfers.
What advantages does HPE Private Cloud AI offer compared to exclusive reliance on public cloud AI?
It provides data control, less dependence on external jurisdictions, and a turnkey integration aimed at production. In sensitive scenarios, it facilitates governance, auditing, and close-to-operation AI deployments.
What does applying multi-agent AI to network operations entail?
It involves coordinating multiple specialized agents (e.g., anomaly detection, traffic optimization, incident response, and capacity planning) working together to perform complex tasks with less human intervention and faster operational speeds.
How does data sovereignty relate to compliance and jurisdictional risks in cloud services?
When data is processed or stored outside the country, it may be subject to other regulations and access requirements. Keeping data “onshore” reduces this exposure and simplifies governance, especially in regulated industries.
via: hpe