On 20 July 2026, e& UAE and Core42 — the G42 sovereign cloud and AI infrastructure company — announced Sovereign AI Compute, positioned by both partners as the UAE's first sovereign GPU compute platform for enterprise AI, with data residency inside the country, zero egress fees and an OpEx-only commercial model.
What e& UAE and Core42 have launched
Sovereign AI Compute is a joint offering that lets UAE-based enterprises and government entities run production AI workloads on GPU infrastructure without moving their data outside national borders. Instead of buying accelerators and standing up their own AI stack, customers consume compute as a service. The operational contour reads like a checklist:
- Data residency inside the UAE — training data, model artefacts and inference traffic stay within the country's jurisdiction.
- Heterogeneous GPU accelerators supplied by Core42, so workloads can be matched to the accelerator that fits, rather than to whatever hardware happens to be in stock.
- Zero egress fees, removing the traffic-out tax that quietly inflates AI bills on hyperscalers.
- OpEx-only model — no upfront capex, on-demand consumption, scale up and down as workloads move.
- Core42 Delivery Services — implementation, integration and platform expertise from the Core42 side of the deal.
- Carrier-grade connectivity from e& UAE — high-bandwidth, low-latency links stitched into the offering.
Alongside the platform, e& UAE is launching GPU Connect — a commercial bundle that packages GPU access together with premium connectivity in a single contract. In effect, buyers get the compute and the network to feed it as one line item, rather than assembling the two separately.
Who it's for: banks, fintech, government, retail, large enterprises
The partners position the platform for organisations that cannot casually send sensitive data to another jurisdiction. In the UAE, that group is not narrow.
Banking and fintech. Institutions supervised by CBUAE, DFSA and VARA have data residency requirements covering customer, transaction and KYC data. Running fraud models, credit scoring or portfolio analytics on infrastructure that stays inside the country becomes an easier compliance conversation than routing the same workloads through an offshore region.
Government and public sector. Federal and emirate-level agencies handling citizen data, regulatory data and public services are the natural buyers for a sovereign stack — particularly for AI-assisted services where the data feeding the model is the sensitive part.
International HQs and regional operations. Multinationals running Middle East, North Africa and South Asia operations from Dubai or Abu Dhabi get a way to keep regional data under UAE jurisdiction while still tapping frontier-scale GPUs — the kind of workload that used to force teams into either a Frankfurt region or a US region by default.
Healthcare, energy, industry and retail. Verticals with either regulated data (medical records, industrial telemetry) or high-volume operational data (retail transactions, supply-chain events, manufacturing sensors) get a platform on which to build sector-specific AI without pulling everything into a hyperscaler contract.
"AI is becoming a key part of how organisations operate. Through our collaboration with Core42, we are enabling UAE enterprises and government entities a simpler, faster and more secure way to access sovereign GPU infrastructure," said Esam Mahmoud, Senior Vice President SMB Sales and Marketing at e& UAE.
The wider agreement between the two companies also covers the Signature National Disaster Recovery Cloud from Core42, aimed at government agencies and operators of critical-sector infrastructure. In other words, a resilience layer sits next to the AI compute layer inside the same partnership.
Regulatory and strategic context
The launch lands on top of a regulatory frame that has been tightening for several years. The UAE's Personal Data Protection Law — Federal Decree-Law 45/2021 — codifies how personal data must be handled and transferred. Sector regulators layer on top of that: CBUAE for banks, DFSA for the DIFC financial ecosystem, VARA for virtual assets. Each of them has data residency requirements inside its perimeter. A sovereign GPU platform maps neatly onto that picture.
It also fits the country's National AI Strategy 2031, which points the UAE at leadership in applied AI rather than at being a downstream consumer of it. Domestic GPU capacity is one of the missing ingredients if the ambition is to have AI built and trained locally, not rented from abroad.
The strategic backdrop for Core42 is the sequence of deals that turned G42 into a hub for frontier compute in the Gulf. In April 2024, Microsoft announced a $1.5 billion investment in G42. In May 2025, G42 and OpenAI were named among the partners behind Stargate UAE. Sovereign AI Compute is the commercial edge of that stack — the point where the underlying capacity meets an enterprise procurement form.
"AI deployment at scale demands infrastructure that can operate securely, compliantly, and with the resilience required for real-world execution," said Jaafar Al Hashmi, VP of Government Relations and Market Development at Core42.
Several practical questions were not addressed in the announcement: published pricing per GPU-hour, which specific accelerator families sit behind the "heterogeneous" label, SLA terms for regulated workloads, and how Sovereign AI Compute will interact commercially with hyperscaler regions already operating inside the UAE. Those will most likely surface when the first named customers start talking publicly.
Based on the joint press release by e& UAE and Core42 distributed via Zawya, and reporting by Compare The Cloud, TradeArabia and TahawulTech.
