The AI-Native Telco: STC-Humain JV, Airtel's leadership shift, O2 Telefónica-Capgemini, and a live 5G slice for emergency services
Telcos are moving from pilots to production in AI and high-capacity compute. This week's moves show where budgets are going: infrastructure, leadership, enterprise AI services, and mission-critical connectivity.
STC's center3 and Humain set up a JV for AI datacentres
STC Group's datacentre and international unit, center3, is creating a joint venture with Humain to build AI-focused facilities in Saudi Arabia. The plan targets up to 1 GW of AI workload capacity, with an initial tranche of up to 250 MW.
The JV combines center3's scale and regional connectivity with Humain's mandate to deliver end-to-end AI infrastructure. The goal: dense compute, high availability, and low-latency routes for large AI workloads across the region.
Why it matters for strategy: scale economics favor early movers with land, energy, and network proximity. Expect co-location demand from hyperscalers and national programs, plus rising interest from enterprises seeking sovereign AI options.
Bharti Airtel confirms CEO transition for 2026
Shashwat Sharma will become managing director and CEO of Bharti Airtel on 1 January 2026. Gopal Vittal moves to executive vice chairman to drive cross-group digital, technology, network strategy, procurement, talent, and long-range planning.
Sharma has served as CEO designate since late 2024 and has worked alongside Vittal over the past year. Airtel ended October with 393.7 million mobile connections in India (33.6% share), behind Reliance Jio's 484.7 million (41.4%).
What to watch: continuity plus renewed growth focus. Expect tighter execution on digital services, 5G monetisation, and cost discipline as India's twin-market dynamic (Airtel/Jio) moves into its next phase.
O2 Telefónica expands AI-native push with Capgemini
O2 Telefónica deepened its relationship with Capgemini to accelerate its AI strategy. Priorities include faster prototyping, iterative deployment across business units, stronger customer experience, network optimisation, and internal efficiency.
The operator also plans a portfolio of AI services for business customers, including AI-based cybersecurity. This signals a shift from "AI inside the telco" to "AI as a product" for enterprise clients.
Implication: expect tighter MLOps, model governance, and reusable components shared across care, network, and enterprise lines-reducing time-to-value and duplication.
MásOrange and Ericsson deliver a live 5G SA slice for Madrid public safety
MásOrange, working with Ericsson, is providing a dedicated 5G standalone network slice for Madrid's public and emergency services. The slice supports secure, reliable, low-latency communications for scenarios ranging from wildfires to natural disasters.
This is a concrete example of network slicing in production with clear performance and priority guarantees. It strengthens the case for premium-grade slices in sectors like transport, utilities, and healthcare.
Executive takeaways
- AI infrastructure moves are getting larger: plan now for land, energy, thermal design, and backhaul. Treat GPU supply, grid contracts, and sovereign data policies as board-level issues.
- Leadership transitions are timing windows: align product, pricing, and capital plans under the new operating rhythm before 2026 budgeting locks in.
- Monetise AI two ways: internal productivity (care, network, back office) and external offers (security, analytics, co-pilots). Build once, reuse many times across units.
- Network slicing is viable where QoS is non-negotiable: prioritize sectors with clear SLAs and procurement cycles (public safety, logistics, industry 4.0).
- Governance matters: standardise data access, model risk, and observability early to avoid fragmentation and compliance rework later.
Next steps
- Run a 90-day plan for AI use cases: pick three with measurable ROI, secure data pipelines, and define production handoff criteria.
- Pressure-test DC expansion assumptions: latency targets, interconnect density, and energy cost curves over 5-7 years.
- Create an enterprise AI catalog: package cybersecurity, analytics, and automation use cases with pricing, SLAs, and reference architectures.
If you're building your leadership team's AI fluency, a curated set of executive-friendly AI strategy courses can help. Explore options by role at Complete AI Training.
Your membership also unlocks:
