This ODNI briefing from 15 May 2007 describes the rationale for ICREACH and the scale of metadata sharing that was envisioned: see the Intercept article The Surveillance Engine: How the NSA Built Its Own Secret Google, 25 August 2014.

This slide from an undated NSA presentation outlines the types of metadata the agency accesses and stores: see the book No Place To Hide, 13 May 2014.

This undated NSA document provides a decision tree for the analysis of email addresses, IP addresses, cookies and phone numbers in XKeyScore and other agency tools: see the Intercept article XKEYSCORE: NSA’s Google for the World’s Private Communications, 1 July 2015.

These two slides, taken from an undated NSA presentation, illustrate the growth in the agency’s metadata collection between 1999 and 2007: see the book No Place To Hide, 13 May 2014.

This NSA Point Paper dated 22 February 2006 provides a history of the CRISSCROSS and PROTON information sharing systems that preceded ICREACH: see the Intercept article The Surveillance Engine: How the NSA Built Its Own Secret Google, 25 August 2014.

This memorandum dated 6 July 2005 describes features of the intelligence sharing systems that preceded ICREACH: see the Intercept article The Surveillance Engine: How the NSA Built Its Own Secret Google, 25 August 2014.

This 23 November 2004 decision memorandum from then-NSA Director Keith Alexander gives the initial go-ahead for the agency’s ICREACH system: see the Intercept article The Surveillance Engine: How the NSA Built Its Own Secret Google, 25 August 2014.