TLDRs
- SoftBank is entering Japan’s battery sector to support fast-growing AI infrastructure demand.
- The company is developing zinc-halogen batteries with South Korean startup COSMOS LAB.
- SoftBank aims to power future AI data centers with safer energy storage systems.
- The project could reduce reliance on lithium and China-linked battery supply chains.
SoftBank Corp. is expanding beyond telecommunications and investment management as it moves into large-scale battery production aimed at supporting the rapid growth of artificial intelligence infrastructure in Japan.
The company announced plans to establish a gigawatt-hour-scale battery business designed to meet soaring electricity demand from AI data centers and advanced computing facilities.
The initiative highlights how the global AI race is no longer focused solely on semiconductors and cloud computing. Increasingly, technology firms are also competing for reliable power generation and energy storage solutions capable of supporting massive AI workloads.
SoftBank said the new business will initially focus on battery systems intended for domestic use in Japan, particularly for next-generation data centers expected to handle energy-intensive AI applications. As companies deploy larger AI models and expand computing capacity, electricity demand has emerged as one of the industry’s most pressing challenges.
Zinc Batteries Gain Attention
As part of the project, SoftBank is partnering with South Korean startup COSMOS LAB to develop zinc-halogen batteries. Unlike traditional lithium-ion batteries, the new systems use pure water as an electrolyte, significantly lowering fire risks while also reducing dependence on expensive raw materials such as lithium and cobalt.
The company believes the chemistry could provide a safer and more stable alternative for large-scale storage facilities powering AI infrastructure. Battery safety has become increasingly important as hyperscale data centers consume more electricity and require larger backup storage systems.
SoftBank’s decision to avoid lithium-based technologies also reflects broader concerns surrounding global battery supply chains. Much of the world’s lithium processing and cobalt sourcing remains connected to Chinese supply networks, creating potential geopolitical and pricing risks for companies building long-term energy infrastructure.
The collaboration with COSMOS LAB remains under a non-disclosure agreement, meaning financial terms and production details were not publicly disclosed. However, SoftBank indicated the batteries could eventually be deployed at major data center facilities planned for Sakai City in Osaka Prefecture.
Massive Storage Goals Ahead
In addition to battery development, SoftBank is also working with DeltaX to manufacture energy storage systems using advanced Cell to Pack technology. The design is intended to improve energy density while optimizing storage efficiency for large-scale commercial use.
SoftBank reportedly plans to conduct performance testing before moving into commercial production, which is currently targeted for 2027. The company is evaluating the possibility of reaching battery production capacity of 1 gigawatt-hour annually by 2030.
If successful, the venture could eventually generate annual sales of approximately 100 billion yen, equivalent to roughly $638 million based on current exchange rates.
The project represents another major infrastructure push from SoftBank as the company positions itself at the center of the AI economy. Founder Masayoshi Son has repeatedly emphasized AI as the next transformative technological wave, and the company has increasingly invested in sectors tied to computing power and digital infrastructure.
Energy Becomes AI Battleground
The announcement also reflects a broader shift happening across the technology industry. While companies initially competed to secure access to AI chips and cloud computing capacity, attention is now rapidly turning toward electricity availability and energy storage.
Modern AI data centers consume enormous amounts of power due to the intensive processing required for training and operating large language models. In many regions, public power grids are struggling to expand quickly enough to meet rising demand.
This has pushed technology firms to explore alternative energy strategies, including battery storage, renewable power projects, and private electricity infrastructure. For companies like SoftBank, controlling both computing resources and energy systems could provide a strategic advantage as AI adoption accelerates worldwide.
At the same time, the company faces challenges. Zinc-halogen battery technology is still relatively new compared to lithium-ion systems, and scaling production may prove difficult. Competition from established Chinese battery manufacturers also remains intense due to their dominant global market position and manufacturing scale.
Still, SoftBank’s move signals growing confidence that energy storage will become a critical pillar of the AI era. As AI models continue expanding in complexity and demand, companies capable of securing stable power supplies may gain an important edge in the next phase of technological competition.


