Can APAC telcos outmanoeuvre SIs and hyperscalers?
In the first of a four-part series on how telcos can stay relevant, we explore the power shift that could sideline the region’s telecoms leaders and the moves they can make to shore up their competitive edge
By
Edwin Lin
Published: 22 Sep 2025
In a recent opinion piece, we explored why Asia-Pacific (APAC) telcos must embrace artificial intelligence (AI) to stay competitive. Now, we turn to a deeper challenge: the shifting power dynamics between telcos, hyperscalers and systems integrators (SIs), and what this means for enterprise transformation across the region.
Canalys forecasts that SIs will capture $43bn of telco-related services by 2029. This is not just new revenue – it’s money that telcos risk losing if they fail to evolve, a sign that those failing to adapt risk being relegated to the role of “dumb pipes”.
SIs: the new enterprise architects
SIs are emerging as vital intermediaries between hyperscalers and enterprises, increasingly occupying territory that telcos have struggled to claim. Their advantage is contextual intelligence. SIs understand local regulations, industry-specific workflows and operational realities – capabilities that hyperscalers and traditional telcos often lack.
South Korea: Samsung SDS’s Cello Square platform uses GenAI to detect global logistics risks and optimise the supply chain, transforming how goods move globally.
Singapore: NCS’s S$130m AI initiative delivers strategic frameworks and industry-specific roadmaps across the Asia-Pacific, enabling enterprise-grade AI adoption.
SIs are becoming enterprise AI architects as they translate hyperscaler capabilities into tailored solutions and capturing high-margin services telcos have long pursued.
Telcos at a crossroads
Since the days of leased lines and multiprotocol label switching (MPLS) networks, telcos have been at the heart of enterprise connectivity. But in the AI era, the economics have shifted.
Connectivity is commoditised, being priced per megabit, easily swapped and increasingly bundled into hyperscaler or SI-led solutions. Legacy networks, once a telco’s crown jewel, now hinder its ability to compete in a software-defined future.
Telstra made early moves into the cloud, recognising the need to diversify. But like many telcos, it struggled to compete with the scale of hyperscalers and the specialisation of SIs. The result: limited traction in enterprise AI.
Hyperscalers dominate cloud platforms. SIs orchestrate enterprise transformation. Telcos, unless they adapt, risk being relegated to passive infrastructure providers in a software-defined world
Singtel, through NCS, provides enterprise-grade IT and AI services while its subsidiary Nxera develops next-generation, AI-optimised datacentres. This allows Singtel to enable enterprise AI adoption without directly competing with hyperscalers.
Telcos must build or partner their way into enterprise AI value chains or risk being squeezed out.
One striking example of telco-led innovation comes from China Telecom. Its Xirang platform hosts a 177-billion-parameter large language model distributed across 500 km of telecommunications infrastructure. For comparison, GPT-3.5 had 175 billion parameters.
Unlike centralised hyperscaler models, Xirang integrates AI computing directly into the telecommunications fabric. This enables low-latency inference across distributed locations while maintaining data sovereignty.
Through this AI integration, more than 160 AI applications have been deployed across 21 key sectors, including industrial manufacturing, transport, smart cities and agriculture, delivering measurable gains in efficiency, cost reduction and quality improvement for enterprises.
Xirang shows how telcos can deliver enterprise capabilities that hyperscalers cannot easily replicate.
Stand still or step up
Across APAC, telcos are being squeezed from both sides. Hyperscalers dominate cloud platforms. SIs orchestrate enterprise transformation. Telcos, unless they adapt, risk being relegated to passive infrastructure providers in a software-defined world.
To stay relevant, telcos must move beyond connectivity and reassert their role in the enterprise AI value chain. In doing so, they face a stark choice: evolve into enterprise enablers or risk being left behind. Three roles are beginning to emerge:
Lead in vertical domains: Use trusted customer relationships and industry knowledge to deliver AI-enabled solutions that directly address sector-specific challenges.
Enable through infrastructure: Build AI-ready networks and platforms like Singtel’s Nxera or China Telecom’s Xirang that provide the latency, compliance and sovereignty advantages that enterprises cannot get from hyperscalers alone.
Orchestrate partnerships: Collaborate with hyperscalers and SIs to co-create value, positioning telcos as the glue between global platforms and local enterprises.
These roles are not mutually exclusive. The telcos that thrive will blend them, leading in domains where they have depth, enabling through infrastructure differentiation and orchestrating ecosystems for reach.
In part two of this series, we will explore how SIs can translate hyperscaler capabilities into business outcomes, embedding themselves in enterprise strategy and capturing the high-margin services that telcos once owned.
Edwin Lin is principal consultant at Omdia, part of Informa TechTarget
Read more on Telecoms networks and broadband communications