Nuclear fusion reactors gain real-time AI shield to tackle plasma collapse risk

Researchers are embedding machine learning into tokamak control systems to spot early tearing modes before they trigger plasma collapse.

The AI models can automatically adjust magnetic fields in real time to suppress instabilities and reduce disruption risk.

This matters because stable plasma control is essential for continuous fusion power generation, especially in larger reactors like ITER.

The work could help make future high-power tokamaks more reliable and closer to practical energy production.