The ranked leadership in this data set is concentrated: five organisations hold 65.4% of the 26 records in scope, and the top ten hold 92.3%. The group is a mix of established fusion-research institutes and materials specialists rather than a single dominant filer, and the ranking extends to 15 companies total, so it should be described as the ranked leaders rather than a top-50 or top-100 list. Anyone assessing freedom to operate should look at the specific claims of the top five before assuming the space is closed.

Filings climbed from zero in 2017 to a peak of 9 in 2023, then dropped to 1 by 2024, a -50% change across the 2021-2024 span. That drop should not be read as the field cooling, because publication typically lags filing by around 18 months, meaning 2025 and 2026 figures in any data set are still incomplete. A fairer read is that 2023 was the busiest year recorded so far, and the true trajectory for 2025-2026 will only become clear once those years finish publishing.

Every record in this scope sits under the G21B fusion-reactor classification by construction of the search, but a smaller share also touches coating and surface deposition, layered composites, alloys, and, more narrowly, welding, scaffolding-related structures and sealing components. Coating technology is the largest secondary category at 11.5% of the 26 records. That drop from 100% to low double digits shows that most of the claim activity centres on the core fusion-reactor concept itself rather than on the surrounding materials and joining techniques.

The technology composition data points to the supporting IPC classes — coating and surface deposition, alloys, layered composites, welding, scaffolding and sealing — each of which covers a small fraction of the 26 records rather than the near-universal coverage of the core fusion-reactor class. A first claim built around a specific joining, coating or composite-integration technique for a plasma-facing surface is likely to face less prior art than one aimed at the limiter or divertor concept itself. Confirming this requires a deeper claim-level search rather than relying on the class counts alone.

The United States receives the largest number of filings among the receiving offices recorded here, with China, Canada, Europe, the United Kingdom and South Korea each contributing a smaller number. This suggests that enforcement risk and freedom-to-operate diligence should prioritise US filings first, then extend to the other offices depending on where commercial deployment or manufacturing is planned. The spread is narrow enough that a global filing strategy in this field is still the exception rather than the norm.