A small cluster of academic research institutions holds the bulk of the foundational claims, with the leading assignee alone accounting for 685 of the 1,402 records in this dataset. The strongest co-assignee pairing shares 604 families, showing that the broadest early claims on CRISPR enzymes, systems and screen methods were filed jointly across a small group of institutions rather than by a single company. Companies working in this space should expect to license or design around this joint-filed core rather than find one clean owner to negotiate with.

Filing peaked in 2018 at 203 records and has since settled onto a lower plateau: the tracked comparison shows 135 records in 2021 against 118 in 2024, a 13% decline over that span rather than a sharp fall. Counts for 2025 and 2026 are still incomplete because publication typically lags actual filing by about 18 months, so the recent years should not be read as evidence of a slowdown. The more accurate read is a maturing core claim set with steady, lower-volume continued filing.

Bioinformatics and readout-analysis claims (IPC class G16B) cover only 4.3% of the 1,402 records in scope, far behind the 71.1% carrying core molecular-engineering claims under C12N. This suggests the analysis and decoding layer — scoring guide assignment accuracy, converting pooled screen data to spatial or in situ formats, and automating screen scalability — carries comparatively less claim density than the underlying perturbation system itself. Teams building software or assay methods for screen readout may find more open claim space here than in the enzyme or delivery-system layer.