
In 2025, AI became the center of gravity for the energy industry. As tech companies commit to over $1 trillion to build out computing infrastructure, utilities are planning to spend another $1 trillion to upgrade the grid. Driven by record spending, support from the federal government, and frenzied competition among LLMs, AI data centers are getting sited, financed, and built at a record pace.
Faced with the speed-to-power imperative, the energy industry has largely taken an “all of the above” approach to development, pursuing batteries, fuel cells, behind-the-meter gas, geothermal, nuclear, and distributed capacity. Solutions are getting more sophisticated, but there is still no uniform blueprint for building at gigawatt scale.
What began as a single urgent question — “How do we accommodate all of this load?” — is evolving into a more coordinated effort to reshape how clean energy is developed, financed and integrated at scale. But the rapid expansion is also colliding with a set of hard constraints:
These pressures reflect a deeper market recalibration. Three years after ChatGPT triggered a construction boom and fueled expectations of a decade-long electricity supercycle, the industry is facing a set of new challenges. Today, skepticism is rising about the durability of AI’s trajectory, the financial stability of key players, and the assumption that ever-larger models will continue to justify ever-larger power demands. The clean energy sector is entering a new phase, one that demands clearer signals, tighter sequencing, and smarter risk management.
Over two days at Transition-AI 2026, attendees will:
Join us for two days in San Francisco as we bring developers, utilities, regulators, and hyperscalers into the same room to align on what’s real, what’s possible, and what can get built that is economically viable and sustainable.