Battery manufacturer CATL has set out its vision for the role of energy storage in the artificial intelligence (AI) era, arguing that the rapid expansion of AI-driven computing will depend on more flexible and scalable power systems.
Speaking at the Siemens Tech Summit China 2026, CATL chief manufacturing officer Ni Jun said the growth of artificial intelligence is increasingly constrained by energy infrastructure, particularly as data centres evolve to handle more intensive and variable workloads.
“At the end of AI is an energy system,” Ni said during a panel discussion alongside Siemens CEO Roland Busch, Alibaba chairman Joseph Tsai and Unitree Robotics CEO Wang Xingxing. “If you don’t have a sustainable energy solution, AI will be limited, because it consumes so much power and energy.”
Ni added that conventional energy systems are not yet equipped to support the requirements of artificial intelligence-native data centres, which demand different power performance characteristics, including rapid response to fluctuating loads and sustained peak capacity.
The discussion reflects a broader shift in how energy storage is positioned within digital infrastructure. As artificial intelligence computing scales, particularly in data centres, energy storage systems are increasingly seen as core infrastructure rather than ancillary support.
CATL is positioning itself to capture this shift by integrating artificial intelligence across its own operations, from battery design through to manufacturing and field performance monitoring. According to Ni, battery production requires extremely high precision, with small variations at micron level affecting overall performance.
AI-driven tools across the value chain
To address this, the company is deploying artificial intelligence-driven tools across the value chain, including design-stage modelling, production-line optimisation and post-deployment data analysis. CATL is working with Siemens to implement digital manufacturing systems that link engineering, production and operational data.
The company said it is also applying the technology to accelerate battery research and development. Using high-throughput computing and large datasets, CATL can simulate material behaviour and screen potential chemistries more rapidly than traditional trial-and-error approaches. Its internal design platform, trained on more than 100,000 design cases and hundreds of terabytes of data, is used to predict performance metrics such as cycle life, energy density and thermal stability.
Beyond R&D, CATL is developing so-called “self-evolving” manufacturing systems through its Next-Generation Smart Line concept. These systems use artificial intelligence to monitor production processes in real time, detect defects and adjust operating parameters automatically, with the aim of improving efficiency and reducing downtime.
The company is also targeting the growing market for energy storage in artificial intelligence data centres. As computing demand rises, operators are facing increasing challenges in managing electricity consumption, particularly given the volatility associated with AI workloads.
CATL said it has partnered with infrastructure providers, including SenseTime, to deploy energy storage systems designed to stabilise power supply and improve efficiency in high-demand computing environments. These systems are intended to provide flexible capacity that can respond to rapid changes in load while supporting overall grid stability.
A recent report by Morgan Stanley highlighted the same trend, identifying energy storage as a key enabler of artificial intelligence growth. The report noted that rising demand for flexible power is likely to accelerate deployment of storage systems across data centres, utilities and grid infrastructure.
Morgan Stanley identified CATL as a potential beneficiary of this shift, citing its scale in battery manufacturing and its expansion into integrated energy storage systems. The bank also pointed to the company’s development of sodium-ion batteries as a possible route to reducing the cost of energy storage, particularly at scale.
CATL has already commercialised sodium-ion technology for certain applications and is preparing for broader deployment in both mobility and stationary storage. At higher production volumes, the company expects the technology to achieve cost parity with lithium iron phosphate (LFP) batteries and potentially deliver further cost reductions.
In parallel, CATL is developing full-system energy storage solutions that combine battery cells, system integration and energy management software. The company said recent projects have demonstrated reductions in both equipment footprint and operating costs, reflecting a shift towards more integrated approaches to energy infrastructure.
Industry speakers at the summit emphasised that artificial intelligence’s long-term impact will be most significant in industrial applications rather than consumer-facing tools. Busch described artificial intelligence as a technology comparable in its impact to electricity, while Tsai highlighted China’s industrial scale as a key advantage in developing and deploying AI-driven manufacturing systems.
Photo: CATL’s Yibin plant (©CATL)
