KUALA LUMPUR (July 30): MBSB Research believes the next phase of digital-infrastructure growth will be determined less by land availability alone and more by the sustainable availability of power, water and connectivity. "At the base of the AI stack, utilities represent both the principal enablers and the binding constraints on further expansion," it said in a note on Thursday. The house said Tenaga Nasional Bhd (KL:TENAGA) remains the most direct beneficiary of rising electricity demand, grid reinforcement, transmission investment and new data centre connections. Meanwhile, as the country transitions to more sustainable and resilient energy systems, companies like YTL Power International Bhd (KL:YTL) continue to stand out as it shifts its focus towards developing green energy projects such as its Green Data Centre Park in Kulai, and upcoming 100MW solar farm in Kelantan. With the launch of the National AI Action Plan 2026-2030 (AI Nation 2030) by the Ministry of Digital, key beneficiaries include companies within the construction, utilities, and renewable energy sectors, according to MBSB. AI Nation 2030 aims to transform Malaysia into “a value-driven AI nation that uses AI to advance inclusive growth, sustainability and the well-being of the rakyat”. To achieve this, the plan outlines two main components — Impact Engines and Foundational Enablers — which focuses on identifying where AI should be deployed to solve sector-specific problems in areas such as healthcare, smart cities, public services, manufacturing, financial services, telecommunications, power utilities, oil and gas services, agriculture, plantations and MSMEs. “Rather than treating AI infrastructure, skills and regulation as separate policy agendas, it links them directly to sector demand,” the research house said. Within the construction sector, players like Sunway Construction Group Bhd (KL:SUNCON) and IJM Corporation Bhd (KL:IJM) are to remain the clearest near-term beneficiaries as data centre contracts aid in order-book replenishment and progress