TechCrunch Disrupt 2026 convenes power infrastructure leaders amid AI energy crisis

AI compute demand is colliding with grid limitations

Electricity has become the primary constraint on AI infrastructure expansion. According to TechCrunch AI, the October 13–15 conference in San Francisco will assemble executives from nuclear fusion developers, grid operators, and data center providers to examine how surging computational requirements are outpacing available power generation and transmission capacity. The event signals that energy—not silicon, not software—now dominates infrastructure conversations within the AI industry.

Fusion vendors taking center stage

Commonwealth Fusion Systems Chief Science Officer Brandon Sorbom and Helion CEO David Kirtley will jointly discuss pathways to commercializing nuclear fusion at grid scale, according to TechCrunch AI. The session emphasizes both technical breakthroughs and remaining engineering obstacles, positioning fusion as a potential long-term solution to data center power demands. The pairing reflects industry consensus that traditional generation sources may be insufficient for projected AI load growth over the next decade.

From Twilio founder to fusion executive

Inertia CEO Jeff Lawson, who founded cloud communications platform Twilio, will participate in a fireside conversation addressing his transition to building fusion infrastructure. TechCrunch AI reports Lawson will discuss how organizational lessons from scaling a software company inform approach to physics-based engineering timelines and recruitment challenges inherent in fusion development.

Grid modernization and utility competition

A panel featuring Drew Baglino (Founder & CEO of Heron Power), Apoorv Bhargava (CEO and Co-founder of WeaveGrid), and other participants will explore infrastructure upgrades needed to match demand growth, according to TechCrunch AI. The discussion examines where private capital is flowing, the role of utility partnerships versus startup disruption, and technical standards for grid flexibility in an electrification-dependent economy.

Data center operators racing for capacity

Sara Spangelo (President & Co-Founder of Ambrosia Energy) and Bill Thayer (SVP, Head of Datacenter Solutions at Bloom Energy) will address how operators are securing power allocation and expanding electrical delivery before supply becomes the binding constraint on AI deployment speed, per TechCrunch AI.

Why This Matters

Energy infrastructure has transitioned from a back-office operations concern to a strategic business constraint. AI teams and investors require visibility into electricity availability, grid geography, and long-term generation capacity to plan deployment timelines. Organizations building AI applications now need to factor in power procurement timelines—sometimes spanning years—alongside compute provisioning. The convergence of fusion startups, established energy vendors, and data center operators at this conference reflects a reshuffling of who controls AI’s expansion path: it is no longer determined by chip availability alone, but by whoever can guarantee reliable, abundant electricity at scale.