Litt on Warrants for Searching American Communications: Either Misleading or Terrifying
At a hearing Wednesday, members of the Privacy & Civil Liberties Oversight Board asked intelligence official Robert Litt a crucial question: If the sweeping general warrants authorized by the FISA Amendments Act are only supposed to be used for “targeting” foreigners for surveillance, shouldn’t a judicial warrant be necessary before NSA can intentionally dig through its massive database of intercepts for Americans’ communications? Otherwise, after all, such “backdoor searches”—currently allowed under NSA guidelines—seem a dangerous loophole that enables an end-run around the rules that would require court approval to directly target an American’s communications for interception.
Litt’s answer was either extremely misleading or extremely disturbing. He told the oversight board that the number of annual queries to the intercept database was “considerably larger” than the few hundred analysts currently run against NSA’s vast archive of telephony metadata records. That would make the “operational burden” of a warrant requirement utterly impractical, Litt asserted, and that the Foreign Intelligence Surveillance Court “would be extremely unhappy if they were required to approve every such query.”
Now, it’s possible that Litt was talking about the total number of queries analysts run against the database of intercepted communications as they sift through it for nuggets of foreign intelligence. No doubt that number is very large indeed. But it’s also utterly irrelevant to the question PCLOB was asking. Nobody, after all, is suggesting that a warrant be required for every query of NSA’s databases—including queries for topical keywords or “selectors” associated with known foreign intelligence targets. The question, rather, was whether a warrant should be required for the subset of those queries involving the name or e-mail address of a particular U.S. person—the very query terms that the government would be forbidden from using as selectors to task interception without first obtaining a particularized, probable cause warrant. If Litt was answering that question by alluding to the total number of queries, then his answer had little bearing on what the PCLOB was trying to discover, and would vastly overstate the practical burden of such a requirement—seriously misleading overseers about the feasibility of a proposed civil liberties safeguard. Litt ought to correct the record if that is what he meant.
What would be hugely more disturbing, however, is if Litt really was giving an answer pertinent to the question he was asked. In that case, he would be representing that NSA runs “considerably more” than a few hundred annual queries for the names and e-mail addresses of specific U.S. persons, against a database of private communications gathered via general warrants—an authority justified on the premise that it is “targeted” exclusively at non-Americans located outside the United States. That would suggest that the blanket surveillance authority created by §702 of the FISA Amendments Act is precisely what civil libertarians feared: A Trojan Horse mechanism for spying on Americans using the pretext of “foreign targeting.”
In short, either added safeguards on NSA’s use of the §702 database are far more feasible than Litt led the PCLOB to believe, or the authority is being used in a way that circumvents constitutional and statutory protections for Americans’ communications on a chilling scale. Litt should clarify which it is—and then Congress should hasten to reform §702 accordingly.