Eight Out Of Ten APAC CIOs To Rely On Edge Services For AI Workloads: IDC

As generative AI becomes essential to business operations, companies are rethinking infrastructure models, according to a new IDC research paper commissioned by Akamai. According to the research paper titled 'The Edge Evolution: Powering Success from Core to Edge,' Asia-Pacific companies are realising that centralised cloud architecture alone is unable to meet the increased demands of scale, speed, and compliance and that's where edge services are necessary to help businesses stay competitive and compliant, and to be ready for AI deployments. 

According to the IDC Worldwide Edge Spending Guide—Forecast, 2025, public cloud services at the edge will grow at a compound annual growth rate of 17% through 2028, with the total spending projected to reach $29 billion by 2028. In addition, in the latest research paper, IDC predicts that by 2027, 80% of CIOs will turn to edge services from cloud providers to meet the performance and compliance demands of AI inferencing. 

Key Points From The Report

As gen AI moves from pilots to execution, enterprises across APAC are confronting the limits of legacy infrastructure. Around 31% of organisations surveyed in the region have already deployed gen AI applications into production. Meanwhile, 64% of organisations are in the testing or pilot phase, testing gen AI across both customer-facing and internal use cases.

However, this rapid momentum is exposing serious gaps in existing cloud architectures:

Complexity Of Multicloud: Around 49% of enterprises struggle to manage multicloud environments due to inconsistent tools, fragmented data management, and challenges in maintaining up-to-date systems across platforms.

Compliance Trap: About 50% of the top 1,000 organisations in Asia-Pacific will struggle with divergent regulatory changes and rapidly evolving compliance standards, and this will challenge their ability to adapt to market conditions and drive AI innovation. 

Bill Shock: Around 24% of organisations identify unpredictable rising cloud costs as a key challenge in their gen AI strategies.

Performance Bottlenecks: Traditional hub-and-spoke cloud models introduce latency that undercuts the performance of real-time AI applications, making them unsuitable for production-scale gen AI workloads.  

Daphne Chung, Research Director at IDC Asia-Pacific, said, “Gen AI is shifting from experimentation to enterprise-wide deployment. As a result, organisations are rethinking how and where their infrastructure operates. Edge strategies are no longer theoretical—they’re being actively implemented to meet real-world demands for intelligence, compliance, and scale.”

Key Findings For APAC

India expands edge infrastructure to meet gen AI demand and manage costs: With 82% of enterprises conducting initial testing of gen AI and 16% leveraging gen AI in production, India is building out edge capabilities in tier 2 and 3 cities. Around 91% of gen AI adopters rely on public cloud IaaS, but cost concerns and skills gaps are pushing demand for affordable, AI-ready infrastructure.

China scales gen AI with edge and public cloud dominance: Almost 37% of enterprises have gen AI in production and 61% are testing, while 96% rely on public cloud IaaS. Edge IT investment is accelerating to support remote operations, disconnected environments, and industry-specific use cases.

Japan accelerates AI infrastructure despite digital maturity gap: While only 38% of Japanese enterprises have gen AI in production, 84% believe gen AI has already disrupted or will disrupt their businesses in the next 18 months, and 98% plan to run AI workloads on public cloud IaaS for training and inferencing workloads. Edge use cases like AI, IoT, and operational support for cloud disconnection are driving infrastructure upgrades.

ASEAN embraces gen AI with edge-first strategies beyond capital hubs: Almost 91% of ASEAN enterprises expect gen AI disruption within 18 months, with 16% having introduced gen AI applications into the production environment and 84% in the initial testing phase. Around 96% are adopting public cloud IaaS for AI workloads, while edge investment is rising to support remote operations and data control.

According to the report, to stay ahead, enterprises must modernise infrastructure across cloud and edge, aligning deployments with specific workload needs. Securing data through Zero Trust frameworks and continuous compliance is essential, as is ensuring interoperability to avoid vendor lock-in. By tapping into ecosystem partners, businesses can accelerate AI deployment and scale faster, smarter, and with greater flexibility.