Glossary
DMCA
The US Digital Millennium Copyright Act, known for its notice-and-takedown process and its rules against removing copyright management information. Both feature in disputes over scraped training data.
Two parts matter for training data: Section 512’s notice-and-takedown process for hosted content, and Section 1202’s prohibition on removing copyright management information such as bylines, credits, and embedded metadata. Several lawsuits over scraped datasets include CMI-removal claims. The DMCA is US law; other jurisdictions run parallel notice regimes.
Why it matters
Keeping credits and metadata intact in a dataset is a compliance requirement, not a courtesy.