Artificial intelligence adoption across Asia Pacific, Japan and Greater China is entering a more decisive phase. What began as experimentation is now maturing into scaled, production-grade deployments, bringing with it new priorities around autonomy, governance and national control over AI infrastructure.
That was the message from Dell Technologies’ regional outlook, shared by John Roese, the company’s Global CTO and Chief AI Officer, alongside Peter Marrs, President for Asia Pacific Japan and Greater China. Together, they outlined how enterprises and governments in the region are reshaping their AI strategies as the technology moves deeper into core operations.
From pilots to production
According to Dell, the conversation around AI has fundamentally shifted. Organisations are no longer testing isolated proofs of concept. Instead, they are asking how to deploy AI systems at scale and how to measure their real impact on productivity, efficiency and business outcomes.
Marrs noted that customers across the region are now focused on operational AI rather than experimentation. This transition, he said, reflects growing confidence in the technology and a clearer understanding of where AI can deliver value.
One example comes from Malaysia, where Sandisk is using Dell infrastructure to power smart manufacturing and advanced product design. The result is a factory operation that now runs at up to 95% “lights-out” automation, a signal of how far AI-enabled industrial systems have progressed.
The rise of agentic AI
A major theme in Dell’s outlook is the accelerating demand for agentic AI, systems designed to handle ongoing, multi-step tasks with a degree of autonomy, rather than acting as simple assistants.
Marrs pointed to Zoho in India as a case study. The company is developing enterprise AI systems that are contextual, multimodal and privacy-first, running on Dell infrastructure. These systems are designed to operate continuously within business workflows, rather than responding to one-off prompts.
Roese described this shift as the beginning of an “autonomous agent era.” In this phase, AI systems start to manage processes end-to-end, reshaping how work gets done inside organisations.
As companies move further along this path, Roese expects the impact to exceed expectations, not by replacing human work, but by amplifying it. Autonomous agents, he said, will quietly take on operational complexity, allowing people to focus on higher-value tasks.
Rebuilding the AI factory
As AI moves into production, Dell argues that enterprises must rethink how they design and operate what Roese calls “AI factories”, the integrated environments where data, models and infrastructure come together repeatedly and reliably.
Resilience is becoming a core requirement. Roese emphasised that AI systems cannot be treated separately from cyber recovery, data protection and system continuity. As AI becomes embedded in day-to-day operations, downtime or data loss carries far greater consequences.
Dell positions itself at the intersection of these needs, combining AI infrastructure with cyber resilience, vaulting and recovery systems to ensure enterprises can stay operational even under stress.
Governance moves to the centre
If agentic AI defines the technical direction of the next phase, governance defines its guardrails.
Roese said that the speed and scale of AI deployment have introduced volatility, forcing organisations to place greater emphasis on control, accountability and compliance. In his view, governance is no longer a constraint that slows innovation, but a prerequisite for sustaining it.
Dell had previously predicted that “agentic” would dominate the AI conversation in 2025. Now, Roese argues, governance will become just as central. Without clear rules, prioritisation and oversight, even the most advanced AI systems will struggle to deliver lasting value.
Sovereign AI gains momentum
Alongside enterprise adoption, Dell sees rapid growth in sovereign AI initiatives across the region. Governments are investing in national AI infrastructure, data policies and local ecosystems designed to keep sensitive data within borders and support domestic innovation.
Marrs said many countries already have foundational frameworks in place. The next step is scaling and refining these systems to support broader economic and industrial goals.
He highlighted Dell’s work with partners such as Macquarie Data Centres in Australia and NAVER Cloud in South Korea, where the focus is on building secure, locally governed infrastructure for what Dell describes as trusted AI innovation.
Roese believes sovereign AI will become a much larger part of the global AI economy than current forecasts suggest, driven by the universal need for reliable infrastructure and locally controlled data platforms.
A regional ecosystem approach
Dell’s outlook also emphasised collaboration as a defining factor in Asia’s AI trajectory. Progress, Marrs said, depends on close coordination between technology providers, governments and academic institutions.
This ecosystem-driven approach is reflected in Dell’s APJ AI Innovation Hub, which brings together infrastructure, talent and partners to support skill development and real-world deployment.
For Dell, the region’s AI future is not just about faster models or bigger systems. It’s about building the foundations, technical, regulatory and organisational, that allow AI to scale responsibly.
