As the Washington Post points out, this opinion, which details many instances in which the NSA flat out lied to the court, lends some credence to statements made by presiding judge Reggie Walton, who claimed the court was limited to making decisions based on information the NSA provided. This opinion appears to detail the NSA setting up its own complicit court system, intentionally misleading it in order to continue its surveillance programs unabated.
[...snip...]
The leaks keep coming and keep pointing to the same conclusion: the NSA has acted as a law unto itself. And all the while it continues to point at its "overseers," which include Congress (which has been lied to directly by the agency when not having information withheld from it by the leaders of the House Intelligence Committee) and the FISA Court (which has been lied to directly and is hampered by its reliance on the NSA's data and narratives -- which pretty much just means more lying).

And despite all this evidence that the NSA's "oversight" is nearly completely compromised, the defenders, including those within the agency, continue to insist the system is working the way it should. In their eyes, maybe it is.

Why?  A little something called "presumption of regularity," which I learned about reading Marcy Wheeler this morning:

Footnote 32 reveals how, after NSA did a review of the communications the FISC ultimately found to violate the Fourth Amendment, the FISC caught it in downplaying the number of affected communications. After it sent the NSA back to new analysis, the problem grew from 2,000 to 10,000 a year to 48,000 to 56,000 a year. I guess the FISC found, like I have, that you can't trust the biggest math organization in the world to do basic math.

Yet in spite of the fact that this opinion lists three substantial misrepresentations the NSA had made in recent history and caught the NSA in bad math, here's how it decided it could trust the government's assurances that it didn't use this abusive communication to target non-targeted people.

   Therefore, the Court has no reason to believe that NSA, by acquiring the Internet transactions containing multiple communications, is targeting anyone other than the user of the selected tasked selector. See United States v. Chemical Found., Inc., 272 U.S. 1, 14-15 (1926) ("The presumption of regularity supports the official acts of public officers, and, in the absence of clear evidence to the contrary, courts presume that they have properly discharged their official duties.").

I'm not surprised FISC invoked this (especially not surprised that John Bates, who can be very deferential, did). It is the law.

But (as the case of Adnan Latif showed) we keep extending the presumption of regularity to the government in spite of abundant evidence we shouldn't.

It's that last point that prompts the real question: if what we've seen so far isn't clear evidence that NSA shouldn't be accorded the presumption of regularity, what would clearer evidence look like?  How bad would it have to get before that line is considered to be crossed and the presumption withdrawn?