The Five Year Parade of Internet Dragnet Violations
Monday’s document release provided mounting evidence that when the hospital confrontation “heroes” moved the Internet dragnet they had deemed to be illegal under the auspices of the FISA Court, neither they, nor Judge Colleen Kollar-Kotelly believed it was legally sound. But they traded those truly crummy legal claims to bring the program under court oversight. Since then, boosters of the scheme have claimed the oversight serves to eliminate violations quickly.
We already knew that’s not true.
Still, Monday’s release — particularly this John Bates opinion written around July 2010 — makes that even more clear. After Kollar-Kotelly sacrificed judicial wisdom for court oversight on July 14, 2004, the government continued breaking the court’s rules for five years, until Reggie Walton shut the program down, sometime in fall 2009.
First, let’s lay out the dates. I’ve done a rough timeline below, based on the known start-date (July 14, 2004) and the rough end point with John Bates’ opinion (around July 2010). The bulk of the other dates impose the timeline laid out in the Bates opinion on a few known dates taken from the phone dragnet production (plus, the geniuses at ODNI not only left the date of the June 22 Internet dragnet order in its URL (CLEANED101.%20Order%20and%20Supplemental%20Order%20%286-22-09%29-sealed.pdf), but it’s the same document as the June 22 phone dragnet order, which has different redactions but most dates intact — see the three bolded entries below).
As you’ll see, there were two known violations in the Internet dragnet before the before the discoveries of the problems started in earnest in 2009. That’s not that big a deal — there was at least one phone violation before 2009 too, except in the case of the Internet dragnet, NSA overcollected from the very start.
The examination of the Internet dragnet started in response to the first phone dragnet disclosures in January 2009 (with the change in Administration, it should be remembered). Reggie Walton told NSA to see if the Internet dragnet had the same compliance problems as the phone dragnet did.
From that point until June 2009, the discoveries seemed to work in parallel (the NSA was working on End-to-End reports for both programs at the same time, and they share some common databases). But with the discovery that both dragnet programs were sharing information freely with other agencies, it became clear the violations were much worse on the Internet dragnet side, with reports going out with US person information that did not even remotely comply with minimization requirements.
Then sometime after that — and after Walton issued what would be the last Internet dragnet order for a year (that was sometime after June 22, 2009) — NSA discovered they had been receiving “metadata” far outside the permitted scope, which surely included content. Note this may have happened around the same time as NSA reported that one phone provider had overproduced (including international data in addition to domestic, I think) on July 9, 2009, so I wonder if they were only then reviewing returned data on receipt.
In any case, it was around that time that NSA “discovered” the Internet metadata program had never ever been in compliance. From Bates:
Notwithstanding this and many similar prior representations [made on the summer 2009 reauthorization] there in fact had been systemic overcollection since [redacted]. On [redacted] the government provided written notice of yet another form of substantial non-compliance discovered by NSA OGC on [redacted] this time involving the acquisition of information beyond the [redacted] authorized categories.
[snip]
This overcollection, which had occurred continuously since the initial authorization in [redacted] included the acquisition of [long redaction]. [my emphasis]
Never.
If my math is correct, the application the NSA withdrew was submitted not long after September 20. There are briefings for the Intelligence Committees that likely alerted them to the scale of the Internet dragnet problems around that time. But as of October 5, some of the most assertive House Judiciary members seem to have had no idea about the problems with the Internet dragnet. If they found out about it with the notice to Congress on December 17, 2009, it explains why the PATRIOT Act reauthorization process stalled.
There’s one more very important thing in this timeline. You’ll see below that almost at exactly the same time as NSA “realized” it had never complied with program requirements, it started a pilot project that would be rolled out on January 3, 2011, analyzing metadata with no special protections for US persons or limit for use only on counterterrorism.
Specifically, these new procedures permit contact chaining, and other analysis, from and through any selector, irrespective of nationality or location, in order to follow or discover valid foreign intelligence targets. (Formerly analysts were required to determine whether or not selectors were associated with US communicants.)
[snip]
In the second place it enables large-scale graph analysis on very large sets of communications metadata vwithout having to check foreignness of every node or address in the graph. Analysts in S2 have used this to great benefit over the past year and a half under a pilot program. [emphasis original]
In other words, at the moment they were coming clean with the FISC that they had never ever complied with the PR/TT orders, they were beginning the pilot project that would move metadata collection overseas, under EO 12333. (This document goes back to this NYT story on social network analysis.)
So much for the notion that putting all this under court oversight would accomplish a damn thing. All it did was degrade the law and provide NSA cover until they developed the technology to do all this overseas.
Update, 11/22: More dates added to timeline.
Update, 11/26: More dates added to timeline. Read more →