du Tech launches National HyperAI sovereign inference platform in UAE

Du Tech's National HyperAI bundles sovereign GPU compute, model serving and agentic workflows for UAE government and enterprise customers.

A brightly lit modern data center hallway features two rows of glass-fronted server racks under overhead lights, leading to large windows displaying a green outdoor landscape.

Du, the UAE-based telecoms and digital services operator, has launched National HyperAI through its du Tech sub-brand, a sovereign AI inference and agentic platform targeting government ministries, large enterprises and SMEs across the United Arab Emirates. The platform is available immediately and brings together GPU as a Service, Model as a Service, and autonomous agent orchestration under a single commercial agreement.

The launch is positioned by the company as a direct response to the UAE's national AI strategy, which calls for the country to become a leading AI-driven economy. Chief ICT Officer Jasim Al Awadi said du Tech is the only provider in the UAE that can offer "complete ownership of the data centres, national connectivity, sovereign inference compute, and an end-to-end delivery journey under one roof." The platform is built on du Tech's National Hypercloud, which the company says is certified by the UAE Cyber Security Council.

What the platform includes

National HyperAI runs across five du-owned data centres in the UAE, utilising liquid-cooled hyperscale infrastructure, bare-metal NVIDIA GPU compute, MLOps orchestration, and national fibre alongside 5.5G connectivity. The company says the architecture allows organisations to connect existing workloads, sovereign databases, and enterprise applications to high-performance AI inference without transferring sensitive data beyond national borders.

By vertically integrating the compute, connectivity and software layers, du Tech claims to remove the multi-vendor complexity that affects competitors relying on leased infrastructure. Customer data, model weights and compute capacity are stated to remain entirely within UAE jurisdiction, addressing data-residency obligations under UAE law. Du did not disclose pricing, specific GPU capacity figures, or the names of any launch customers in its release. The reference to technology ecosystem partners is unspecified, and no benchmarks for inference throughput or latency were provided.

Market context and competitive positioning

Sovereign AI infrastructure is a rapidly expanding category across the Gulf Cooperation Council, with Saudi Arabia, Qatar and the UAE all investing in nationally controlled compute to reduce dependence on US hyperscalers. G42, the Abu Dhabi-based AI holding company, has established relationships with Microsoft and operated large-scale GPU infrastructure in the region for several years, making it du's most prominent domestic competitor in the enterprise AI-infrastructure segment.

Globally, managed sovereign AI platforms are being developed by telecoms groups and national cloud operators in France, Germany, Singapore and India, often in response to government procurement requirements that mandate data residency. The EU AI Act and analogous national frameworks in the Gulf explicitly require regulated organisations to maintain accountability and audit trails over AI systems, which strengthens the commercial case for single-operator sovereign stacks.

Regulatory and standards read-across

UAE federal AI regulations, alongside the UAE Cyber Security Council's cloud certification programme, create a compliance incentive for public-sector buyers to consolidate workloads with a domestically owned operator rather than a foreign hyperscaler operating a local zone. Du's claim to UAE Cyber Security Council certification for its underlying Hypercloud is commercially significant in this context, though the company has not stated which certification level or standard applies. Enterprises in regulated sectors, including banking, healthcare and defence-adjacent government functions, will want to verify specific compliance scope before procurement decisions.

The agentic AI layer is worth watching as a differentiator. Autonomous agent frameworks are moving rapidly from development tools to production infrastructure, and embedding them inside a sovereign stack alongside model serving and GPU compute could lower the barrier to production deployment for UAE organisations that have so far been constrained to controlled pilots.