stc and Humain to build AI-focused data centers in Saudi Arabia with up to 1GW capacity
Saudi Telecom Co. (stc) and Public Investment Fund-backed Future Artificial Intelligence Co. (Humain) have agreed to form a joint venture to develop and operate AI-dedicated data centers in the Kingdom. Humain will hold 51 percent and stc 49 percent, with the build executed through stc's subsidiary center3.
The agreement runs for six years under a memorandum of understanding. The design targets up to 1 gigawatt of total power, starting at 250 megawatts based on customer contracts.
Deal snapshot
- Structure: JV between Humain (51%) and stc (49%) via center3
- Scope: AI-specific data centers with high-capacity, low-latency infrastructure
- Power: Up to 1GW, initial 250MW ramped by demand
- Term: Six-year MoU with phased build-out
- Disclosure: Related-party classification as Humain is a PIF subsidiary and PIF holds a stake in stc
Why this matters for engineering teams
This move strengthens Saudi Arabia's AI compute footprint and regional connectivity, which can cut latency for training and inference at scale. Expect options that better support data residency, enterprise integration, and capacity planning for GPU-intensive workloads.
For teams deploying models in the GCC, this could reduce cross-border hops, simplify compliance, and open procurement paths for large, reserved capacity.
Technical footprint and ramp
The program is designed for high-capacity, low-latency infrastructure, with center3's regional network as a backbone. The power envelope-1GW with a 250MW starting block-signals intent to serve large-scale AI workloads as contracts land.
Build-out will be sequenced. Capacity and timelines will track customer commitments and vendor availability.
Financials and governance
The financial impact is expected to be positive but is not yet quantifiable, given dependencies on project plans and investment requirements. The related-party note reflects PIF's involvement on both sides of the JV.
Policy context
The initiative supports Saudi Arabia's Vision 2030 targets on digital infrastructure and AI-driven growth. It also aligns with national goals to climb global rankings in open data and AI publications by 2030.
Saudi AI momentum
Earlier this month, Saudi Arabia ranked fifth worldwide-and first in the Arab region-for AI sector growth, according to the Global AI Index. That trend signals expanding demand for compute, talent, and production-grade MLOps across the Kingdom.
What builders should watch
- Location details, availability windows, and latency targets between zones and carriers
- Capacity reservation terms for AI clusters and onboarding queues
- Data residency controls, audit pathways, and cross-border traffic policies
- Network fabrics and interconnect options for hybrid and multi-cloud
- SLAs around uptime, thermals, and high-density racks
- Contractual flexibility for scaling from pilot to production
Action items for IT and development teams
- Map AI workloads by tier: training, fine-tuning, batch inference, real-time inference
- Define data governance for Saudi residency and cross-region replication
- Benchmark expected GPU/accelerator demand and plan reserved capacity
- Design network paths to center3 PoPs and validate interconnect performance
- Align FinOps with cluster usage patterns (utilization targets, queueing, spot vs. reserved)
- Harden MLOps: environment reproducibility, model registry, and audit trails
Upskilling for the new capacity wave
If you're building or scaling AI systems in the region, now's the time to level up skills across data engineering, MLOps, and secure deployment. Curated paths for technical roles can speed up onboarding and help teams make use of new compute as it comes online.
Your membership also unlocks:
