More than a few folks have commented on the irony of Sen. Dianne Feinstein—perhaps the most vigorous defender of NSA’s controversial surveillance programs—flipping into high dudgeon mode when surveillance is targeted at her powerful allies, such as Germany’s Angela Merkel, or her own Intelligence Committee staffers. At a surface level, though, Feinstein could reasonably argue that no hypocrisy is involved here: Specifically seeking to monitor the activity of congressional overseers, she might claim, bucks democratic accountability in an inherently objectionable way that broad data collection does not, provided only appropriate uses are made of the data collected.  Which is fair enough. Even for those already disturbed by, say, the NSA’s bulk telephony program, it would surely be more disturbing to learn that analysts had specifically queried that bulk database for the phone records of prominent political figures. It was such monitoring, after all—not merely whimsical violation of politically disengaged Americans’ privacy—that lay at the heart of the intelligence scandals of the 1970s.

But worry not, Alanis fans, ironies remain. For one, while Feinstein did allude to staffers’ “work product” being stored on a network drive accessed by CIA, some of the most revealing information about the committee’s activity would have been in the form of metadata—or as Feinstein likes to say, “just metadata,” analysis of which the senator has insisted is not really surveillance at all. Access logs on the staff computers would, after all, give a detailed picture of which documents were accessed when, and for how long, by committee investigators—and multiple staffers accessing the same document in rapid succession, especially if it coincided with a flurry of communications between them, would surely be a signal that something was afoot. Even if we assume the agency foreswore scrutiny of search terms entered by staff, these could often be reverse engineered from the size of the query string and volume of results returned. Such logs would tip the CIA off to the direction of the committee’s inquiry, and also serve as a guide to which documents  had not been accessed, and therefore could be surreptitiously removed without raising red flags.

Were such monitoring limited to metadata, it would not, pace Feinstein, constitute a violation of the Fourth Amendment under current doctrine, since these records are stored on the CIA’s own “third party” systems. Nor would it matter that the CIA had explicitly agreed to respect the integrity of the committee’s investigation on those borrowed machines. As the Supreme Court has repeatedly made clear, individuals “assume the risk” that even the most ironclad-seeming contractual guarantees of privacy might be betrayed, thus waiving their “reasonable expectation of privacy” however vigorously they might endeavor to secure it by such agreements.

The larger contradiction, however, is that the incredible scale of NSA surveillance, which makes really granular monitoring effectively impossible, implicitly requires its defenders to fall back on the claim that the intelligence community can be trusted to use its authorities only or legitimate purposes—and to cooperate with overseers in identifying and correcting episodic violations of the rules. Yet what Feinstein describes is an agency determined to stymie and “intimidate” its overseers in order to avoid public embarrassment. House Minority Leader Nancy Pelosi even suggested that Feinstein—chair, remember, of the Senate Intelligence Committee—might pay a “price” for her fiery speech, because when the CIA is criticized, “they come after you and they don’t always tell the truth.” In light of such casual remarks, the overseers start to sound a bit like the adults in the classic Twilight Zone episode “It’s a Good Life,” reassuring us that the intelligence community has done a real good thing—and hoping they won’t be sent into the corn.