EDGAR (SEC Filings)
EDGAR (Electronic Data Gathering, Analysis, and Retrieval) is where US public companies file their disclosures with the SEC. It covers decades of filings: yearly and quarterly financial reports, material event notices, insider trades, registration statements, and shareholder proxy materials.
You can reach the data through full-text search, daily and quarterly index files, or the SEC's submission and company facts APIs, which return structured JSON. Filings themselves come as HTML, plain text, and XBRL, a tagged XML format that carries the numbers in financial statements as machine-readable fields.
Everything on EDGAR is a US Government work and sits in the public domain, so you can reuse it freely. It is a common source for RAG systems that answer questions about company financials, filings history, and regulatory compliance.
Related sources
Cambridge Law Corpus
A research dataset of more than 250,000 UK court cases, mostly from the 21st century but with some reaching back to the 16th century. Built by the University of Cambridge for legal natural language processing work, it is available under restricted access for research use.
EUR-Lex
The official portal for European Union law. It provides EU treaties, legislation, case law, and legislative proposals in 24 languages, with bulk download available. A dependable source for multilingual legal and regulatory RAG across the EU.
Free Law Project / CourtListener
The largest open collection of US court opinions, oral argument recordings, and judicial financial disclosures, run by the nonprofit Free Law Project. Its RECAP Archive holds hundreds of millions of federal court docket entries, and journalists at the Wall Street Journal and ProPublica have used it for investigations. A solid base for US legal RAG.
Legislation.gov.uk
The official UK government legislation database, holding all UK Acts of Parliament and statutory instruments. It is available as XML and linked data, so you can load the full statute book into a RAG system rather than scraping individual pages.