IBM puts sovereign infrastructure and agent governance at the centre of Asia-Pacific AI plan

IBM has set out a blueprint for Asia-Pacific organisations to embed artificial intelligence more widely across their operations by 2030, placing digital sovereignty, data readiness, and the governance of AI agents at the centre of the plan.

Presented at IBM THINK on Tour Singapore, the framework addresses the move from isolated AI experiments towards coordinated systems in which models and agents interact with enterprise data, applications, and operational processes. Guest of Honour Josephine Teo, Singapore’s Minister for Digital Development and Information, also attended the event.

The announcement combines IBM’s regional adoption strategy with additions to its enterprise AI and hybrid cloud portfolio. It also draws on projects involving Singapore’s defence and healthcare sectors, as well as a collaboration intended to support locally controlled AI infrastructure across Asia.

Sovereignty becomes part of the AI operating model

IBM shares that organisations will need to treat data residency, regulatory compliance, and control over infrastructure as core elements of AI deployment. These requirements become more complex when businesses operate across several markets, use a mixture of public and private cloud environments, or allow autonomous agents to access data and initiate actions across multiple systems.

According to research cited by IBM, 68% of executives view data residency and sovereignty as a challenge. The company also reported that 72% of Asia-Pacific chief executives are willing to make strategic decisions based on AI, creating pressure to expand adoption while retaining oversight of where data is stored and how systems are governed.

Hans Dekkers, General Manager of IBM Asia Pacific, noted that Asia Pacific is uniquely positioned to be a leader in the race to AI-first enterprises. He added, “The region is already navigating the realities of fragmented regulations, data sovereignty requirements, and complex hybrid environments at scale. By treating this complexity as a strategic advantage, and building AI that is sovereign by design, organisations here are setting the global benchmark for how trusted AI is governed and scaled.”

The blueprint links these sovereignty requirements with the operational foundations needed to deploy AI across an organisation. This includes preparing enterprise data for AI use, setting controls around models and agents, and establishing infrastructure that can support workloads across different environments without giving up local governance or operational independence.

IBM cited further research showing that 79% of executives globally expect AI to contribute significantly to revenue by 2030. About 80% are increasing investment in agentic AI, while spending on AI agents is expected to nearly triple by 2027.

Dekkers described the requirements in broader terms, stating, “The enterprise AI race is already underway, and by 2030, the leaders will be organisations that embed AI across every aspect of their business to transform operations, accelerate innovation, and create lasting competitive advantage. Success will depend on three essentials, trusted AI, AI-ready data, and an open, hybrid, sovereign digital foundation built for scale. Organisations that get these fundamentals right will move beyond experimentation to operationalise AI with confidence, unlocking the speed, agility, and resilience needed to lead in the AI economy.”

Product additions focus on orchestration and operational control

IBM presented several enterprise AI and hybrid cloud capabilities alongside the blueprint. The next generation of IBM watsonx Orchestrate is intended to coordinate multiple AI agents, addressing the need to manage workflows that involve several specialised systems rather than a single model or assistant.

IBM Confluent is designed to provide AI applications with real-time data, while IBM Concert focuses on intelligent operations. IBM Sovereign Core is aimed at organisations seeking greater operational independence over their technology environments. IBM Bob, meanwhile, is described as an agentic AI software development partner built for enterprise teams.

The portfolio reflects the different layers involved in moving agentic AI into production. Agents need access to current business data, a mechanism for coordinating tasks, controls governing what actions they can perform, and infrastructure that allows organisations to manage security and compliance obligations across markets.

These requirements also create dependencies between AI deployment and existing enterprise architecture. An orchestration layer cannot compensate for inaccurate data, unclear permissions, or fragmented systems, while sovereign infrastructure still needs to connect with the applications and data sources that support day-to-day operations.

Regional projects move the plan into operational settings

In Singapore, the Digital and Intelligence Service and Defence Science and Technology Agency are working with IBM to build quantum computing expertise and explore how the technology could eventually support complex mission-planning requirements.

Military Expert 7 Guo Jinghua, Commander of the SAF C4 and Digitalisation Command, highlighted the strategic value of the collaboration, stating, “Quantum computing has the potential to transform how complex mission planning problems are solved, and we believe it is important to engage early to understand both its opportunities and limitations. Through this collaboration with IBM, the DIS and DSTA are building quantum expertise, strengthening partnerships with industry and exploring how emerging technologies can be applied to future operational challenges. By investing in our people and capabilities today, we will be better prepared to harness quantum technologies as they mature.”

The work remains focused on developing expertise and assessing potential applications while the technology matures, rather than presenting quantum computing as an immediate replacement for existing operational systems.

IBM is also working with Visionbay.ai on a Build-Operate-Localise model intended to establish locally controlled AI infrastructure, support enterprise adoption, and develop regional AI capabilities. Visionbay.ai is backed by Hon Hai Technology Group, also known as Foxconn.

Neo Yao, CEO of Visionbay.ai, underscored the broader ecosystem impact, stating, “IBM and Visionbay.ai will help nations establish AI as a new digital industry, enabling local ownership of critical infrastructure, accelerating enterprise adoption, developing AI talent, and creating economic opportunities across the ecosystem. Our Build-Operate-Localise model represents the next evolution of AI infrastructure. Together with IBM, we are creating the foundation for trusted AI adoption across Asia.”

A separate healthcare project at Changi General Hospital combines AI, mixed reality, and clinical expertise for paediatric mass casualty incident preparedness. The hospital is using watsonx Orchestrate for agentic workflow orchestration and performance tracking, with the system intended to provide real-time scenario intelligence, generate objective performance insights, and support more consistent decision-making across healthcare teams.

The project is attributed to Adjunct Assistant Professor Jimmy Goh, Senior Consultant at Changi General Hospital’s Department of Emergency Medicine. It illustrates how IBM’s orchestration approach could be applied where several participants, data points, and clinical decisions must be coordinated within a time-sensitive training environment.