Why are there redactions in the case law scans?

Many of the decisions in our case law database link to digital scans of their source. For example, scans of Miranda v. Arizona can be found here.

Many of these scans contain extensive redactions:

These redactions cover up potentially copyrighted editorial content that was in the source material. This could be things like headnotes, summaries, etc. that the publisher of the volume created.

Many of these scans come from the Harvard Caselaw Access Project, which used humans to draw redaction boxes. Other scans come from our scanning initiative, in which we use a machine learning tool called Blackletter to black-out relevant content.

We're on record in the Third Circuit arguing that headnotes are not copyrightable, but we redact out of an abundance of caution.