
CATL, the world's largest EV battery maker, is repositioning itself as an AI infrastructure investor by backing DeepSeek and investing in data centres and power systems. Rather than building AI chips or large language models directly, the company is focusing on the energy layer — providing batteries, power systems, and thermal solutions that AI data centres depend on. This strategy leverages CATL's existing expertise in battery technology while tapping into the high-power demands of AI computing.
Summaries like this, in your inbox every morning.
Sign up free →What happened
CATL is expanding into AI infrastructure through strategic investments in DeepSeek, data centres, and power systems, positioning itself at the energy layer of the AI computing value chain and extending its business beyond electric vehicles.
Why it matters
By investing in AI infrastructure rather than competing directly in chip or model development, CATL is leveraging its existing strength in battery and energy systems to serve a critical bottleneck in AI operations — power and thermal management — where demand from data centre operators is likely to remain intense.
What to watch
The company's success will depend on whether its energy-first strategy can differentiate it from other infrastructure players and generate sufficient returns from the AI sector to offset any slowdown in EV demand.
CATL, the world's largest battery manufacturer for electric vehicles, is reorienting its growth strategy toward AI infrastructure. The company is making a series of strategic moves to establish itself at the energy layer of the AI computing value chain — the backbone systems that power and cool large-scale AI operations. These investments include a stake in DeepSeek, a leading AI model developer, as well as direct investments in data centres and power systems. Rather than attempting to build competitive AI chips, processing units, or large language models from scratch, CATL is leveraging its decades of experience in battery design, energy storage, and power management to address one of the most acute bottlenecks in AI infrastructure: reliable, high-capacity power delivery and thermal management. Data centres running advanced AI models consume enormous amounts of electricity and generate intense heat, creating continuous demand for sophisticated battery systems, power conversion equipment, and cooling solutions — areas where CATL has established manufacturing scale and technical depth. This energy-first strategy extends CATL's addressable market well beyond the EV sector, which has shown signs of maturation in some regions. By focusing on infrastructure rather than competing head-to-head in model development or chip design, CATL is applying its core competitive advantages to a sector where demand is projected to remain high as AI adoption accelerates globally.
CATL's move into AI infrastructure represents a strategic pivot from its dominance in the EV battery market. Rather than chasing AI chip or model development — areas where the company lacks deep technical heritage — CATL is applying its core competency in battery technology and power systems to a new market. The energy layer of AI infrastructure has emerged as a critical constraint: data centres running large language models consume enormous amounts of power and generate significant heat, creating sustained demand for advanced battery systems, power management, and thermal solutions. By investing in DeepSeek alongside data centre and power system development, CATL is positioning itself not as a direct AI competitor but as an essential infrastructure provider. This approach allows the company to extend its total addressable market beyond the maturing EV sector while maintaining competitive advantage based on existing manufacturing and supply-chain strength.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack