The genesis of SES AI’s journey lies in the ambitious research undertaken at MIT by its founder, Dr. Qichao Hu. Initially, Hu’s work aimed to address the extreme conditions faced in oil and gas exploration, where sensors operate at temperatures exceeding 120°C (250°F). The goal was to engineer batteries capable of withstanding such harsh environments and offering extended operational life. This led to the development of a solid polymer lithium metal battery, a technology that promised significantly higher energy density compared to conventional lithium-ion batteries prevalent in consumer electronics and EVs. These advanced cells utilize lithium metal for their anode and a polymer electrolyte, enabling a more compact and potent energy storage solution.

The initial venture, then known as Solid Energy, spun out of MIT in 2012, securing its first private investment in 2013. However, the niche market of oil and gas exploration proved insufficient for substantial growth. Recognizing the burgeoning potential of electric vehicles as they entered the mainstream, the company recalibrated its strategy. After optimizing its battery chemistry for lower operating temperatures, SES AI established pilot manufacturing facilities, first in Massachusetts and subsequently in Shanghai, China.

By 2021, the global battery industry was experiencing a boom, with EVs at the forefront of innovation. Major automakers like General Motors, Hyundai, and Honda expressed keen interest in next-generation battery technologies, leading to collaborations with SES AI. The company identified larger vehicles, such as SUVs and trucks, as particularly suitable candidates for their advanced batteries. The rationale was that these heavier vehicles would benefit from lighter, more energy-dense batteries to achieve a practical driving range without an excessive weight penalty.

A further evolution in SES AI’s technological approach occurred in 2022 with the announcement of a shift from lithium metal anodes to silicon anodes. This change was motivated by the potential for simpler and more cost-effective manufacturing processes. However, subsequent developments have introduced new complexities. The growth of the EV market in the United States has experienced a notable slowdown, partly attributed to a reduction in government support, including the expiration of crucial EV tax credits for consumers in late 2025. This market deceleration, particularly for larger electric vehicles, has compelled SES AI to re-evaluate its market focus.

In response to these evolving market dynamics, SES AI is now placing its strategic bets on its AI-powered materials discovery platform, named "Molecular Universe." This platform represents a significant departure from its previous emphasis on physical battery production. The core idea is to leverage AI to accelerate the identification and development of novel battery materials. The company plans to monetize this platform in two primary ways: by licensing its AI software to other battery manufacturers and by developing and selling proprietary battery materials discovered through its AI algorithms.

Molecular Universe has already demonstrated its potential, with SES AI reporting the identification of six new electrolyte materials. One of these is an additive designed to enhance the longevity of batteries utilizing silicon anodes. A persistent challenge with silicon anodes is their tendency to swell during operation, which can lead to physical degradation and compromise charging efficiency. The industry currently relies on materials like fluoroethylene carbonate (FEC) to form a protective, elastic film on the anode, enabling effective charging. However, FEC can degrade at high temperatures, releasing gases that can shorten battery lifespan. The SES AI platform has reportedly identified a compound that mimics FEC’s protective properties without generating these detrimental gases.

Hu emphasizes that SES AI’s extensive history and deep-seated expertise in battery science are critical to the success of its AI platform. He argues that the company’s domain knowledge and the vast dataset accumulated from years of battery development and testing are more valuable than the AI models themselves. By stepping away from the capital-intensive and complex process of physical battery manufacturing, SES AI believes it can achieve faster scalability and revenue generation. "By not actually making the physical battery, we’re actually able to scale and then generate revenue faster," Hu stated.

However, the transition to an AI-driven materials discovery model is not without its skeptics. Kara Rodby, a technical principal at Volta Energy Technologies, a venture capital firm specializing in energy storage, expresses reservations about the immediate impact of AI in materials discovery on the broader battery industry. She notes that while new materials development is theoretically desirable, it may not be the primary bottleneck hindering progress in battery technology. "New materials development, as much as we thought that was what people wanted (and, frankly, it should be what the cell makers want)—I don’t know that that seems to be the real linchpin of the battery industry’s progress," Rodby commented.

Rodby further points to the current investment climate, with investors pulling back and public support waning, making the path forward challenging for many in the battery sector. She questions whether the ability to discover new materials, in isolation, can unlock significant advancements at this juncture. "I don’t know that the ability to discover any new material is going to unlock anything new for the battery industry at this point in time," she concluded.

This strategic pivot by SES AI highlights a broader trend within the battery industry, where innovation is increasingly occurring at the software and materials science levels, rather than solely through incremental improvements in manufacturing. The company’s gamble on AI for materials discovery, while met with some caution, represents a forward-thinking approach to navigating a complex and rapidly evolving technological landscape, potentially reshaping the future of energy storage and its geopolitical implications. The ability to rapidly identify and license novel materials, powered by AI, could offer a more agile and profitable path than the traditional high-volume manufacturing model, especially in an era of shifting market demands and geopolitical considerations in the global supply chain for critical energy technologies.