SES AI's Pivot From Batteries to Software Reveals a Deeper Crisis in Western Energy Storage

Qichao Hu, the CEO of SES AI, has a blunt diagnosis for an industry he helped build. "Almost every Western battery company has either died or is going to die. It's kind of the reality," he says. That kind of candor is rare in a sector that spent the better part of a decade attracting billions in venture capital, government subsidies, and breathless coverage about energy independence. But Hu's pivot away from battery manufacturing toward artificial intelligence software isn't just a corporate strategy shift. It's a signal flare about the structural forces that have made it nearly impossible for Western battery startups to compete on hardware alone.

SES AI, based in Massachusetts, once harbored ambitions of producing lithium-metal batteries at scale, a technology that promised higher energy density than conventional lithium-ion cells. Those ambitions have been significantly scaled back. The company is now redirecting its focus toward AI-driven battery diagnostics and management software, essentially betting that the intelligence layer sitting on top of battery systems is where Western firms can still carve out defensible ground. It's a pragmatic retreat, but it also reflects something more systemic about how the global battery industry has evolved and who controls it.

The story of Western battery startups over the past decade is, with painful regularity, a story of the gap between laboratory promise and factory-floor reality. Companies like Northvolt, which filed for bankruptcy protection in late 2024 after burning through billions, and a string of American SPAC-era darlings that never reached meaningful production volumes, all ran into the same wall: manufacturing batteries at cost-competitive scale requires not just capital but decades of accumulated process knowledge that Chinese firms, led by CATL and BYD, have already internalized.

CATL alone controls roughly 37 percent of the global EV battery market, according to data tracked by SNE Research. Chinese manufacturers benefit from deeply integrated supply chains, lower labor costs, state financing, and the kind of learning-curve advantages that come from producing at volumes Western competitors simply haven't matched. When a startup in Massachusetts or Germany tries to compete on cell chemistry alone, it's entering a race where the other runners have been on the track for years.

This is the structural trap Hu is describing. It isn't defeatism so much as a systems-level observation: the feedback loops that drive down manufacturing costs reward incumbents with scale, and those incumbents are overwhelmingly Chinese. Every quarter a Western startup spends trying to close that gap is a quarter of cash burn against a moving target.

The pivot to AI software is, in one sense, a classic response to being outcompeted on hardware: move up the value chain. SES AI's bet is that battery management, diagnostics, and predictive analytics represent a layer where proprietary data and algorithmic sophistication can create moats that a Shenzhen factory cannot easily replicate. There's logic to this. As EV fleets scale and grid storage deployments multiply, the operational complexity of managing thousands of battery systems creates genuine demand for intelligent software.

But the second-order consequence worth watching is what this retreat means for Western energy security ambitions more broadly. Governments in the United States and Europe have spent years and tens of billions of dollars trying to build domestic battery supply chains, precisely because dependence on Chinese-manufactured cells is seen as a strategic vulnerability. The Inflation Reduction Act alone directed enormous incentives toward domestic battery production. If the companies those policies were designed to nurture are now concluding that manufacturing is unwinnable and pivoting to software, the policy architecture starts to look like it's subsidizing a transition rather than a solution.

There's also a subtler risk. If Western firms cede the hardware layer entirely and compete only on software and services, they become, in effect, dependent on the same Chinese-manufactured cells they're helping to optimize. The intelligence layer and the physical layer are not easily separated over the long run. Whoever controls the hardware tends, eventually, to develop the software expertise too.

Hu's candor is valuable precisely because it cuts through the optimism that has characterized so much of the clean energy investment narrative. The battery industry's consolidation around a handful of Asian manufacturers isn't a temporary imbalance waiting to be corrected by the next breakthrough chemistry. It's a durable structural reality shaped by scale, policy, and time. The more interesting question now is whether Western governments will update their industrial strategies accordingly, or continue funding a hardware race that the startups themselves are quietly exiting.