The rapid pace of technological change shows no signs of slowing down, and the year 2025 will be no different. It is crucial for businesses and digital leaders across industries to stay ahead of these shifts to fully harness the potential of emerging technologies and overcome the challenges they bring, ultimately driving growth. Malaysia is no exception, as its dynamic economy and growing digital landscape demand proactive adaptation to these advancements.

Looking ahead to 2025 and beyond, there are four key trends expected to shape the business and technology landscape in the Asia-Pacific region.

Trend 1: Embracing a hybrid approach to AI deployment, incorporating private AI infrastructure

Much of the initial boom in AI service deployment has been facilitated by the availability of Large Language Models (LLMs) available in the public cloud. However, we are seeing digital enterprises increasingly recognizing that alternative infrastructure approaches might be better suited for some AI workloads – particularly those involving private data. 

Rather than sending user queries and their associated data to be processed by models in the public cloud – an approach we can term ‘Data To The Model’ – many organizations are now using a ‘Model To The Data’ approach. This involves deploying AI models on private computing infrastructure, adjacent to the organization’s private data storage, typically in physical locations which are close to the end-users of the model.  This approach can potentially offer benefits from the perspectives of privacy, speed and cost.

First, privacy.  Many of the most promising AI use cases involve applying the technology to data sets which may be sensitive, private or even regulated – think of fraud protection systems for banks leveraging financial records, or preventative healthcare services depending on medical images.  To keep full control over sensitive data, or to comply with national data sovereignty requirements, some enterprises are now choosing to deploy AI infrastructure privately. The impetus to do this may be stronger for organizations wishing to train a customized model with their own proprietary data. In Malaysia, where sectors like banking, healthcare, and e-commerce are highly regulated, compliance with the Personal Data Protection Act (PDPA) and sector-specific rules is a significant consideration. Deploying private AI infrastructure allows businesses to maintain full control over sensitive data, meet national data sovereignty requirements, and gain a competitive edge.

Speed is another consideration.  The seamless use of AI services is dependent on the latency of the connection between users and the computing infrastructure which houses the service, as well as interconnectivity between ecosystems. This requirement for low latency will only increase as AI services increasingly evolve from short text-based queries, to audio, image, and video-based services. To achieve low latency, organizations can leverage highly distributed colocation data centers as an option for their AI infrastructure deployment – in locations close to their base of end-users. Organizations in Malaysia can capitalize on the country’s growing network of colocation data centers in locations like Cyberjaya and Johor Bahru, allowing AI infrastructure to be deployed closer to users for faster service delivery.

Finally, the cost structure of different AI services needs to be considered. Services at an early stage of development, or services which involve limited amounts of data transfer, might be better served by a model sitting in a public cloud and available on-demand.  For mature services involving large amounts of data transfer between users and the model, the running costs of private infrastructure may be lower. In Malaysia, where industries like manufacturing, fintech, and logistics are seeing increased reliance on AI, hybrid solutions that combine public and private infrastructure can help optimize both performance and costs.

In 2025, we expect to see an increase in the proportion of enterprises deploying hybrid AI infrastructure, giving them the flexibility to leverage both private and public infrastructure depending on these considerations for specific AI services. Malaysia’s strategic position as a growing digital hub and its commitment to innovation through initiatives like the Malaysia Digital Economy Blueprint (MyDigital) will likely accelerate the adoption of these hybrid AI infrastructures.

Trend 2: Strengthening cybersecurity with the growing prowess of AI and quantum

Cyber threats are on the rise in the Asia-Pacific region, with cybersecurity spending expected to reach $36 billion in 2024. This surge is driven by various factors, including the increasing sophistication of cyberattacks leveraging AI and IoT technologies. According to Gartner, trends such as generative AI, unsecure employee behavior, third-party risks, continuous threat exposure and identity-first approaches are exposing new threats to the cybersecurity landscape. Similarly, the proliferation of IoT devices expands the attack surface, making robust security measures more critical than ever.

Quantum computing will become another key accelerating threat to cybersecurity. It presents serious risks to critical elements of today’s public key infrastructure and is expected to be able to break today’s encryption in minutes. In fact, nation-state actors are already harvesting encrypted sensitive data with the intention to decrypt it later when the technology is available, so-called “harvest now, decrypt later” attacks.

To combat these threats, quantum cryptography and generative AI tools are becoming an essential component of cybersecurity strategies for organizations. For example, Quantum Key Distribution as a Service (QaaS), a cloud service enabling internet access to quantum key distribution technologies, offers robust protection for private enterprise networks, ensuring secure communication and data integrity. The gradual integration of quantum technologies into cybersecurity frameworks will be paramount in protecting data from increasingly advanced cryptographic attacks. Quantum Key Distribution will provide unprecedented levels of security, safeguarding sensitive data against these sophisticated cyber threats.

Trend 3: Leveraging edge computing to enhance data sovereignty

Governments’ increasing focus on data sovereignty, coupled with the rise of Internet of Things (IoT), generative AI, and real-time applications, necessitates robust IT infrastructure at the edge. Edge computing allows for localized data processing, reducing transfer risks and ensuring compliance with national data sovereignty laws, which vary widely across Asia-Pacific. Enterprise IT leaders increasingly recognize the need for an edge compute strategy: Gartner report ‘Building an Edge Computing Strategy’found that “by year-end 2026, 70% of large enterprises will have a documented strategy for edge computing, compared to fewer than 10% in 2023.”

In the Asia-Pacific region, several countries have begun implementing stringent data sovereignty policies to protect their citizens’ data. For instance, China’s Cybersecurity Law mandates that data collected within the country must be stored domestically. Similarly, Indonesia’s Government Regulation No. 71 requires electronic system operators to store data locally. These regulations highlight the need for localized data processing solutions like edge computing to facilitate compliance and secure data handling.

Edge computing services are increasingly relied upon to allow businesses and governments to process data closer to the source, achieving faster response times and improved data security. This localized approach will be essential for complying with data sovereignty regulations, particularly in regions with stringent data protection laws. For example, Malaysia’s healthcare sector, projected to reach USD 30 billion by 2027, increasingly relies on low-latency solutions like edge computing to enable telemedicine and real-time diagnostics. These technologies are particularly vital for remote areas, where edge computing can bridge connectivity gaps and ensure that patients have access to timely and accurate medical services, thereby improving overall healthcare outcomes and reducing the strain on urban healthcare facilities.

Trend 4: Advancing business applications with hybrid multicloud

Asia-Pacific is home to 37% of the world’s cloud data centers, with the Asia-Pacific public cloud market expected to grow at a CAGR of 26% until 2026, with expansions planned in markets like Malaysia, Thailand and Vietnam.  Google recently announced a $2 billion investment to establish its first data center and Google Cloud region in Malaysia, supporting local industries and the public sector as part of the country’s digital strategy. Similarly, Amazon Web Services (AWS) launched the AWS Asia Pacific (Malaysia) Region in 2024, marking the largest investment by an international cloud provider in the country to date, with a planned investment of MYR 29.2 billion (USD 6.2 billion) through 2038.

At the same time however, we are seeing many enterprises adopt a hybrid multi-cloud approach, bringing together the agility of multiple public cloud services, with the benefits of private cloud infrastructure. Several factors are driving enterprises towards hybrid multicloud solutions, including GPU scarcity, unpredictable cloud costs, the price-to-performance ratio for predictable workloads, and specific use cases that require data to be completely under the customer’s control.

Hybrid multicloud will continue to be the standard for enterprises seeking to optimize their IT infrastructure, balancing the benefits of public and private clouds.

Stay agile by tapping into digital interconnection services

As we move towards 2025, the Asia-Pacific region is set to experience significant technological advancements. From the development of private AI environments and quantum-enhanced cybersecurity, to the accelerated adoption of edge computing and hybrid multicloud solutions, these trends will shape the future of the digital landscape.

Moreover, with rising sustainability concerns, it’s increasingly crucial for businesses to adopt these advanced technologies in a manner that minimizes environmental impact.

Cheam Tat Inn is Managing Director of Equinix, Malaysia

By admin