The CEO of supply chain giant Flex says infrastructure and the grid needs to scale with AI—or else ‘Forget all the frontier models’

AI is becoming ubiquitous in both professional and personal realms, from chatbots integrated into workflows to sprawling data centers sprouting up in suburban communities. However, Revathi Advaithi, CEO of supply chain giant Flex, warns that America's infrastructure may be ill-prepared to handle the rapid surge of AI technology.

Speaking at Fortune’s annual Most Powerful Women Summit, Advaithi emphasized that while many discussions revolve around semiconductors and frontier models, she remains focused on the critical need for a robust energy infrastructure and grid system. "If you don’t get the energy infrastructure and the grid ready, forget all the frontier models and forget all the semiconductors," she cautioned.

Building a power grid capable of meeting the country’s burgeoning energy demands is a decades-long endeavor, whereas developing a chip typically takes only five years. As AI continues to permeate various aspects of life, the real concern, according to Advaithi, is whether the nation’s power infrastructure is prepared for such a transformative shift.

By 2030, approximately 11.8% of U.S. electricity may be allocated to data centers, as per a 2025 Berkeley study. Advaithi expressed that the country is significantly lagging behind in power infrastructure innovation, and transforming this system will take considerable time. She estimated that the U.S. will still face a power gap within the next decade as AI technologies evolve.

To keep pace with AI’s growth, Advaithi stressed the urgent need to expedite standards and approvals for innovation. However, she acknowledged that significant hurdles exist, including a plethora of approval processes for new power technologies and a lack of urgency and innovation within the data center sector. The Flex CEO highlighted that the slow progress in power infrastructure, marked by safety concerns and regulatory impediments, is detrimental to AI’s burgeoning appetite for electricity.

Advaithi warned that if the current sluggish pace persists, AI’s increasing power demands could outstrip the capacity of existing infrastructure, potentially undermining the very technology it aims to support.