Sites Blocked by Internet Filtering Programs

Edelman Expert Report for
Multnomah County Public Library et al., vs. United States of America, et al.

Overview

In the course of a pending ACLU challenge to the Children's Internet Protection Act (CIPA) (PDF format), I was asked to design and implement systems to identify particular web pages that are blocked by four specific Internet filtering programs but which do not fit within the programs' self-defined categories for blocking. I was also asked to identify and describe the capabilities and flaws of widely-used Internet blocking systems. This web site reflects a portion of this work and a portion of my written report.

My testing focused on Surfcontrol Cyber Patrol 6, N2H2 Internet Filtering 2.0, Secure Computing SmartFilter 3.0, and Websense Enterprise 4.3. My Expert Report (PDF) documents specific programs and sites tested, testing methodology, and results. My subsequent Rebuttal Report (PDF) notes certain addititional shortcomings in blocking programs, as well as certain flaws in the methodology of expert witnesses for the United States. My Supplemental Report (PDF) details the extent to which overblocking continues over time.

Update: The United States District Court for the Eastern District of Pennsylvania issued a decision overturning CIPA. The decision states in part: "Any public library that adheres to CIPA's conditions will necessarily restrict patrons' access to a substantial amount of protected speech in violation of the First Amendment." (May 31, 2002)

Update: Oral argument will be heard in the Supreme Court on March 5. Briefs are available. (February 13, 2003)

Expert Reports

  • Initial Expert Report (October 15, 2001) (see appendices below)
  • Rebuttal Report (November 30, 2001)
  • Supplemental Report (March 13, 2002) (see appendix, Excel spreadsheet, >600KB)
  • Documentation of Specific Pages Blocked

    In total, my research yielded 6777 distinct web page URLs that were blocked by at least one of the filtering programs tested, as configured. Of these pages, council for the plaintiffs designated 395 URLs to be listed in the first file linked below ("Appendix A" to my initial expert report), while the other 6382 URLs are in the second file below ("Appendix B").

  • Subset of blocked pages with page details
    Includes page title (as retrieved by Google and manually verified), page description (as retrieved by Google, if any), Yahoo categorizations (if any), Google categorizations (if any), and filtering program blockages, categories, and dates of testing. Submission to evidence also includes site archives as well as archives of linked pages. (395 URLs total)
  • Additional pages with page details (warning: >3MB)
    Includes page title (as retrieved by Google), page description (as retrieved by Google, if any), Yahoo categorizations (if any), Google categorizations (if any), and filtering program blockages, categories, and dates of testing. Submission to evidence also includes site archives. (6382 URLs total)
  • Pages in Joe Janes Random Sample with page details (warning: >500KB)
    Includes page title (as retrieved by Google), page description (as retrieved by Google, if any), Yahoo categorizations (if any), Google categorizations (if any), and filtering program blockages, categories, and dates of testing. Submission to evidence also includes site archives. (859 URLs total)
  • Plaintiff sites with selected details

    Includes page title, and filtering program blockages, categories, and dates of testing. Submission to evidence also includes site archives as well as archives of linked pages. (12 URLs total)

    Notes

    A protective order limits distribution of certain portions of my expert work. However, redacted versions of my expert report and rebuttal report are available (PDF).

    This page is made available to inform discussion about the features and abilities of Internet filtering software. The data contained here is not intended for use for other purposes, and it should not be used for other purposes without first contacting the author.


    Ben Edelman
    Last Updated: February 12, 2003 - Notify me of major updates and additions to this page.

    This page is hosted on a server operated by the Berkman Center for Internet & Society at Harvard Law School, using space made available to me in my capacity as a Berkman Center affiliate for academic and other scholarly work. The work is my own, and the Berkman Center neither funded it nor expresses a position on its contents.