SoftBank Group and Grab Holdings have announced plans to invest $1.85 bn in artificial intelligence infrastructure across Malaysia, in partnership with Petroleum Sarawak Berhad (PETROS).
The companies signed joint development agreements covering projects in Sarawak and Peninsular Malaysia, with a long-term target of developing at least five gigawatts of AI computing capacity.
- SoftBank Group, Grab and PETROS have signed development agreements for an initial $1.85 bn investment in AI infrastructure across Sarawak and Peninsular Malaysia.
- The projects target at least 5 GW of AI computing capacity over the long term, with additional financing expected from global investors, hyperscalers and project lenders.
- SoftBank will contribute $1 bn, Grab $750 mn and PETROS $100 mn, supporting Malaysia’s ambition to become a regional AI technology hub by 2030.
The proposed investment forms part of a multi-year initiative to expand Southeast Asia’s computing infrastructure and support rising demand for artificial intelligence services.
Once completed, the facilities are expected to rank among the largest AI infrastructure developments in Asia, providing computing resources to governments, multinational companies and smaller businesses.
The projects will be developed in stages, with capital deployment dependent on development milestones, customer agreements and other closing conditions.
SoftBank, Grab and PETROS allocate $1.85 bn to AI projects
The investment will be divided between two major development projects.
SoftBank Group has committed to providing $1 bn in proposed initial capital, while Grab will contribute $750 mn and PETROS will invest $100 mn.
The Sarawak project will receive $1 bn in initial funding, with SoftBank providing the majority of the capital. The Peninsular Malaysia project will receive $850 mn, with Grab serving as its founding shareholder and principal developer.
Malaysia AI infrastructure investment breakdown
| Investor | Sarawak | Peninsular Malaysia | Total |
|---|---|---|---|
| SoftBank Group | $800 mn | $200 mn | $1 bn |
| Grab Holdings | $100 mn | $650 mn | $750 mn |
| PETROS | $100 mn | — | $100 mn |
| Total | $1 bn | $850 mn | $1.85 bn |
The companies intend to attract additional investors and hyperscale cloud customers as the projects advance.
Future development could also be financed through third-party equity and non-recourse project financing, limiting direct financial obligations at the parent-company level.
Sarawak AI infrastructure project targets long-term expansion
The Sarawak project will be jointly developed by SoftBank, Grab and PETROS under an agreement building on their October 8, 2026, framework arrangement.
SoftBank will initially provide $800 mn, representing the largest share of the proposed investment. Grab and PETROS will each contribute $100 mn.
Grab will have an option to increase its ownership position, although SoftBank is expected to remain the majority shareholder.
The partnership combines SoftBank’s AI investment experience, Grab’s regional technology capabilities and PETROS’ knowledge of energy infrastructure in Sarawak.
The companies will work with the Sarawak Government to develop facilities capable of supporting large-scale AI computing operations.
Grab to lead $850 mn AI infrastructure project in Peninsular Malaysia
The second project will focus on Peninsular Malaysia, where Grab will lead development. The proposed initial investment totals $850 mn, comprising $650 mn from Grab and $200 mn from SoftBank.
Grab plans to introduce additional strategic investors as the development expands, diversifying the project’s funding sources.
Anthony Tan, co-founder and CEO of Grab, is expected to oversee both Malaysian AI infrastructure projects and appoint dedicated management teams to handle daily operations. He will continue to lead Grab, and the company’s board has approved his involvement in the developments.
SoftBank has also agreed in principle to a five-year lock-up covering Grab shares held directly by an entity within its group, subject to agreed terms and customary exceptions.
Malaysia targets at least 5 GW of AI computing capacity
The projects are intended to support Malaysia’s National AI Action Plan 2026–2030, which aims to strengthen the country’s position as a regional AI technology hub.
Research by Google, Temasek and Bain & Company indicates that per capita AI usage across Southeast Asia’s six largest economies is approximately one-tenth of the level recorded in the United States.
The research projects that regional AI usage could increase by as much as 60 times by 2030.
SoftBank and Grab expect this growth to create substantial demand for computing infrastructure located within Southeast Asia.
Masayoshi Son, chairman and CEO of SoftBank Group, said Malaysia has the technological capabilities and development ambitions required to support the next phase of regional AI adoption.
Anthony Tan said the projects would help accelerate localized AI development while creating opportunities for Malaysian businesses and workers.
AI data centers to focus on energy efficiency and sustainability
Electricity consumption, water availability and grid capacity are becoming increasingly important considerations in AI infrastructure development.
The Malaysian projects are expected to incorporate high-efficiency systems designed to reduce water consumption and dependence on local electricity grids.
The facilities will operate in accordance with Malaysia’s renewable energy guidelines.
Development will involve cooperation with the federal government and Sarawak authorities to align infrastructure investment with national economic and environmental priorities.
The partners also intend to explore opportunities in advanced manufacturing, engineering and other technology-intensive industries.
These initiatives could support employment in technical occupations while expanding Malaysia’s capabilities in AI infrastructure design, construction and operations.
SoftBank and Grab expand Southeast Asia AI investments
The agreement represents a significant expansion of SoftBank’s and Grab’s investment activities in Southeast Asia.
- For SoftBank, the projects provide an opportunity to develop large-scale computing infrastructure in a region experiencing growing demand for AI services.
- For Grab, the investment extends its involvement in artificial intelligence beyond its existing digital services and technology operations.
The companies expect the infrastructure to support the development of AI applications tailored to Southeast Asian markets, including services for businesses, public institutions and consumers.
However, the full 5 GW capacity target will require substantial additional development and financing beyond the proposed $1.85 bn initial investment. Project execution will depend on securing customers, electricity capacity, infrastructure approvals and long-term funding.
The partnership is expected to contribute to Malaysia’s AI Nation 2030 strategy while strengthening the country’s role in the regional digital economy.
