First Light Fusion has raised £25 million in the first close of its latest funding round, led by East X Ventures and Starmaker One, with significant participation from the UK Atomic Energy Authority. Existing shareholders, including IP Group, also supported the round. The UK-based inertial fusion energy company will use the capital to accelerate commercial development of its FLARE Fusion Energy concept, which aims to provide a simpler, lower-cost pathway to fusion energy. IP Group currently holds a 28.3% undiluted beneficial interest in First Light, with an additional 17.3% held via its managed funds. The investment marks a strategic endorsement of First Light's technology and its relevance to the UK's long-term energy ambitions as it advances towards scalable, reactor-compatible fusion.

The UK government is investing £45 million in Sunrise, an AI-driven supercomputer dedicated to fusion energy research, set to launch in June at the UK Atomic Energy Authority's Culham campus. The system will deliver up to 6.76 exaFLOPS of AI-accelerated modelling performance. Sunrise will combine high-performance computing with physics-informed AI models to run detailed simulations and create digital twins of fusion systems, helping researchers model plasma turbulence, develop reactor materials and advance tritium fuel breeding before conducting physical experiments. The machine incorporates AMD EPYC processors and AMD Instinct GPU accelerators on Dell PowerEdge infrastructure. The supercomputer will support UK fusion initiatives including the LIBRTI programme and the STEP project, a prototype spherical tokamak power plant planned for Nottinghamshire in the 2040s.

CERN and the UK Atomic Energy Authority have developed PipeINEER, a 3.7-centimetre-wide robot designed to autonomously inspect the Large Hadron Collider's 27-kilometre beam pipes. The robot navigates areas inaccessible to humans, where superconducting magnets operate at −271°C under high vacuum conditions. PipeINEER travels up to six kilometres on battery power, capturing images of approximately 2,000 plug-in modules that handle temperature and pressure extremes. An AI model detects deformations in radio frequency components that could obstruct the beamline. When issues are identified, the robot returns to report precise locations, eliminating the need to disassemble pipe sections for manual inspection. The robot was developed by UKAEA's Remote Applications in Challenging Environments robotics centre, applying expertise from fusion energy research to particle physics.