Second Working Thread on Exigent Letter IG Report
It has taken me a while. But I’ve finally gotten through the DOJ IG Report on exigent letters. Page numbers below will be to the PDF page.
Page 14: Footnote 1 notes there are Secret and TS/SCI versions of this report. Keep that in mind as you read the redactions–while it’s probably safe to assume that Feingold and Wyden (who are both on SSCI) have seen the entire report, it’s not clear who else will have seen the entire report.
Page 14: I hadn’t really noticed it before, but the time frame on the first IG Report’s investigation of exigent letters ended on December 16, 2005–the day that Eric Lichtblau and James Risen exposed the illegal wiretap program. That suggests that the use of exigent letters, among other things, may have changed on that date in response to the discovery of the program. Also note that in Fine’s first report on NSLs, he decided to lump 2005 in which the time frame–2002 to 2004–required by statute. This is parallel to what he did with Section 215, suggesting that there were significant changes in 2006 after disclosure of the overall program.
Page 18: Note that the IG Report doesn’t say when the Public Integrity Section declined to prosecute these abuses. I do hope Fine gets asked that question.
Page 24: Notes that most exigent letters issues from April 2003 to March 2006. That latter date suggests they implemented a fix with the PATRIOT revision passed that month.
Page 28: Note the organization of the Communication Exploitation Section (CXS):
- Document Exploitation (which became Digital Media Exploitation on March 26, 2006
- Communication Analysis Unit (the section that issued the exigent letters, and therefore working on communities of interest)
- Electronic Communication Analysis Unit (how does this differ from CAU???) (ECAU)
- Electronic Surveillance Operations and Sharing Unit (EOPS)
Does this suggest the EOPS collected this stuff and the others did network analysis on it?
Page 29: Note the final date for the exigent letter range here is November 13, 2006, which is different from the December 16, 2005 used elsewhere
Page 34: Note how Company A (AT&T per EFF’s math) does something (maybe “analyze” toll records) that the other two providers don’t do (per footnote 26).This is almost certainly the community of interest analysis. This may sugges that by default mean they were working with massive data collection, since it would mean they had access to the signals of their competitors?
There also must be internet analysis in here (which presumably might be the ECAU), which itself would seem to require telecom assistance. So I wonder whether that fully-redacted paragraph describes a contract that does both phone and internet analysis?
Page 35: Does the redaction showing the size of the contract midway down the page appear to be 10 digits? Suggesting the contract would be in the single million range? (That making the digits something like this: $X,XXX,XXX) Though the amount for Company B seems to consist of words, not just numbers.
Page 36: The language about whether companies were able to provide subscriber data or not closely resembles language surrounding Section 215, which was used during some of this time period to get subscriber data (though possibly in larger batches). And note the redacted second half of the first full paragraph on this page says that they were also doing something in addition to giving meta data and subscriber data. And footnote 28, saying that Company A would only provide subscriber data, suggests that that company (AT&T?) was demanding more than one of the others was, legally.