Microsoft and Partners Accelerate Fusion Energy with AI at the Inaugural Fusion Summit
Microsoft Research hosted its first-ever Fusion Summit in March 2025, bringing together global leaders in AI and fusion energy research to explore how AI can help accelerate the quest for sustainable, clean fusion energy, according to a recent blog by the company.
The event highlighted AI’s potential to enhance the development of fusion reactors and address the challenges of this complex energy source.
The summit kicked off with insights from Ashley Llorens, Corporate Vice President of Microsoft Research Accelerator, who shared a vision of using AI to drive sustainability in energy.
Steven Cowley, director of the Princeton Plasma Physics Laboratory (PPPL), delivered a keynote on the scientific and engineering hurdles of fusion reactors and the need for international collaboration. Cowley emphasized the critical role that AI and high-performance computing will play in modeling fusion reactor designs to advance progress.
AI Applications in Fusion Research
At the summit, representatives from DIII-D, the largest fusion facility in North America, demonstrated how AI is already being used to optimize reactor design and operations.
At the peak, DIII-D's representatives, who are the largest fusion complex in North America, showcased how AI is being utilized already to streamline reactor design and operations.
Richard Buttery and Dave Humphreys of General Atomics demonstrated AI-powered innovations like active plasma control to reduce instabilities, density limits from machine learning that will provide safer operations, and AI-based trajectories to prevent harmful tearing modes in the plasma. These technologies focus on enhancing the performance of reactors and providing safer high-density operations.
One major challenge that was mentioned was the creation of the "first wall" of fusion reactors, which must be able to withstand extreme temperatures and particle bombardment.
Zulfi Alam, corporate vice president of Microsoft Quantum, discussed how quantum computing could help address these material challenges, particularly in terms of hydrogen diffusion and finding new materials that can withstand neutron and alpha particle damage.
Broader Impact of AI on Fusion Engineering
Throughout the summit, AI’s broader potential in fusion research was explored in lightning talks from Microsoft Research labs.
Topics ranged from using AI in gaming for plasma control to employing AI in robotics for remote maintenance, physics-informed AI for simulating materials and plasma behavior, and surrogate modeling of fusion’s underlying physics.
Archie Manoharan, director of nuclear engineering for Cloud Operations at Microsoft, emphasized that AI could play a pivotal role in optimizing fusion systems.
Also read: Solis Unveils Hybrid Inverter and Smart AI Assistant at Intersolar 2025
Global Collaboration and Next Steps
A panel featuring experts like Steven Cowley and Microsoft’s Chris Bishop emphasized the vital role of global collaboration and AI in advancing fusion energy.
Microsoft is working with ITER and Princeton Plasma Physics Lab to support fusion ignition and optimize experiments through AI, highlighting AI's growing importance in overcoming fusion’s scientific and regulatory challenges.
