AI-ML service codevelopment

AI services are a key focus of telco-hyperscaler collaboration. Telcos boast large datasets and vertical market reach; hyperscalers have the algorithms and compute. SK Telecom is partnering with AWS to establish an “AI Zone” data center in Uslan, South Korea, to help South Korean organizations develop AI applications locally. Spain’s Telefónica is working with Google Cloud on AI and GenAI initiatives – among other endeavors, training large models on telco data and building an AI stack for communication services and developer platforms. In general, telcos see hyperscalers’ AI-ML capabilities as a way to add intelligence to networks and services such as predictive maintenance, customer analytics, cybersecurity. 

Many telcos now have R&D labs or AI hubs with cloud partners, and these tend to be long-term deals. Among other examples, see Vodafone’s $1.5 billion deal with Microsoft and subsequent $1 billion deal with Google, both billed as 10-year strategic partnerships.

Sovereign clouds as competitive bastions

Working with hyperscalers to deliver high-value IT and computing services to all sorts of businesses is opening up real growth opportunities for telcos. The second major diversification domain for telcos involves helping organizations adopt sovereign clouds – that is, cloud technologies that provide full control of data, infrastructure, and compliance in line with local laws and regulations.

The combination of governmental data-protection regimes such as GDPR and geopolitical realignments driven by emerging nationalistic tendencies have the potential to stoke growth in governmental and corporate sovereign clouds.

A hyperscaler with data centers in dozens of countries – and which is subject to the dictates of the U.S. Cloud Act, which lets U.S. law enforcement demand customer data regardless of where it’s stored – faces regulatory and market skepticism on the sovereign-cloud front. Telcos can leverage their status as trusted homegrown providers. That’s valid whether they team with hyperscalers or develop their own branded sovereign cloud businesses with domestic IT partners.

Regulatory and technological forces will shape how deeply telcos and hyperscalers intertwine. Regulatory pressure looks likely to mount. Data protection, privacy, and national security laws may force operators and cloud providers to localize key functions. We may see mandates that telecom core networks or certain data stay within national borders under local control.

A given telco may well pursue combinations of hyperscaler coopetition and telco-led sovereign cloud development. Take the example of Telus. The Canadian operator signed a 10-year Google Cloud deal in 2021. Then in September 2025, it opened with several non-hyperscaler partners the country’s first fully sovereign AI data center in Quebec, “unlocking advanced AI capabilities for Canadian businesses, researchers and innovators while storing data within our national borders, in a data centre under Canadian control,” as Telus officials put it.

Whether sovereign clouds coalesce into globally competitive platforms or remain fragmented by region remains an open question. Even as new forces pushing data and computing balkanization gain steam, initiatives such as Gaia-X and ED-EDIC in Europe, not to mention good old economies of scale and Metcalfe’s law, will continue to serve as counterweights.

Other important factors will also shape sovereign clouds. Hyperscalers and other global partners can help efficiently and cost-effectively build and operate sovereign clouds. That helps ensure consistency across borders, thereby fostering, rather than hampering, innovation. True data sovereignty is not about technological isolation, but rather the ability for public authorities to verify where data is processed, who operates critical systems, which legal framework applies, and who is accountable at every level, as SAP CEO Christian Klein recently described it.

Either way, telcos are in a great position to take the lead on sovereign clouds and harness them as growth engines.

New telecom business models in telecom don’t require radical departures from what drove past successes. Ubiquitous mobile computing, AI, explosively growth in data volumes, and customer expectations of split-second access are further solidifying telcos’ strategic importance throughout the global economy. Telcos are finding partnerships with hyperscalers to be a great way to move from commodity data transfer into high-value IT and specialized communication services. At the same time, heightened political attention to data and computing sovereignty are stoking interest in sovereign clouds, and telcos everywhere are well positioned to capitalize.