This was crossposted on The American Prospect
After a brief hiatus, legislative reform of the NSA’s bulk collection program appears to be back on track. Thanks to skillful negotiations on the part of Senator Patrick Leahy, Democrat of Vermont, and other cosponsors, the version of the USA Freedom Act that was unveiled in the Senate last week restores many of the protections that House leadership and administration officials stripped out of the House version in secret, last-minute talks. Most notably, the Senate bill clearly would prohibit the bulk collection of Americans’ telephone records and other types of information.
And yet, even if the Senate version becomes law, Americans’ private information will remain vulnerable—under both the domestic programs addressed by the bill and other, much larger, programs nominally targeted at foreigners. As Leahy acknowledged when introducing the bill, much more remains to be done to protect the privacy and civil liberties of law-abiding Americans.
The good news first: The Senate bill would require any collection of business records to be based on a “specific selection term,” such as a name or account, that narrowly limits the scope of collection “to the greatest extent practicable.” The bill includes a non-exhaustive list of selection terms that are deemed too broad, including area codes, zip codes, and names of telecommunications companies. The Foreign Intelligence Surveillance Court (FISA court) would have to approve the selection terms in advance and assess whether the records would be relevant to an authorized investigation. These provisions would not only end the NSA’s bulk collection of telephone records; they would preclude any analogous program for Internet, financial, or credit records.
The bill also leans on the executive branch to be more transparent about surveillance activities. It would require the director of national intelligence to make public either a redacted version or a summary of any significant opinion by the FISA Court. It also would require far more detailed statistical reporting on the use of surveillance authorities. For the first time, the government would publicly report the number of individuals affected by various surveillance programs—including, for most programs, a separate estimate of the number of affected Americans. And the bill would establish a panel of paid privacy advocates who could appear in FISA court proceedings, which currently take place with only government officials present.
In these and many other respects, the bill should lead to a marked improvement over the status quo. But the operative word here is “should.” The bill’s definition of “specific selection term” is necessarily imprecise. Congress would not limit the executive branch to obtaining the records of named suspects, because in some cases the very reason for seeking the records will be to identify the suspect —for instance, where the FBI has a tip about a plot to bomb a particular airliner and seeks to obtain the passenger manifest. Rather than impose an exact but too-strict definition, the bill attempts to focus collection through phrases like “narrowly limit” and a list of terms that would not be narrow enough.
Although the legislative intent behind this approach is clear, the executive branch could exploit the absence of a bright-line restriction to engage in collection that is far broader than necessary, even if it falls short of “bulk” collection. It would not be the first time the executive branch twisted Congress’s words with the FISA court’s blessing.
Under the bill’s requirement to disclose significant FISA court interpretations, the public should know if such a perversion of congressional intent has taken place. But here, too, there is the potential for the executive branch to disregard the spirit of the legislation. The bill allows the executive branch to decide which court opinions meet the bill’s definition of “significant” and how much information may be disclosed consistent with national security. In theory, nothing would prevent the director of national intelligence from releasing an opinion with every sentence but one redacted.
Even assuming faithful implementation, the bill leaves some holes. It specifies no time limits for when intelligence agencies must discard information about Americans that has not been deemed to have any foreign intelligence value. It implicitly accepts the administration’s specious claim that it cannot even estimate the number of Americans whose communications are swept up under a program that targets foreigners’ calls and e-mails. And it allows the FISA Court to determine when, if ever, it wishes to hear from the panel of privacy advocates.
The bill also creates a new telephone records collection program. While the executive branch would not be allowed to collect phone metadata in bulk, it would be entitled to obtain records, not only of suspected terrorists, but of anyone in contact with them – an automatic second “hop.” Phone companies would produce all such records to the government on an ongoing basis, and the resulting database could be kept indefinitely and queried for any purpose. Given that two independent, presidentially-appointed committees concluded the telephone metadata program had little value, it is unclear, at best, why even a scaled-down version of it is needed.
Of course, the perfect should never be the enemy of the good—and on Capitol Hill, where the perfect is generally off the table, the good should not be the enemy of the better. But the opposite is true as well. A bill that makes improvements to the status quo, even significant ones, can backfire if it dissipates the incentive for additional reform.
That risk deserves particular attention here, where so many of the government’s most intrusive surveillance practices have been left to another day. The bill’s architects deliberately deferred substantive reform of Section 702 of the FISA Amendments Act, a 2008 law that allows the NSA to collect the content of Americans’ international communications without a warrant. Nor does the bill tackle the overseas collection of communications under Executive Order 12333, which operates with no involvement of the FISA court and little oversight by Congress. If bulk collection is the tip of the surveillance iceberg, alarming because of its visibility, these other programs are the looming underside.
Reaching consensus on how to reform these massive, poorly understood programs will be challenging, to say the least. Leahy and his co-sponsors were right not to hold other reforms hostage. But if the bill becomes law, lawmakers and the public must not betray its promise by assuming that the balance between our liberties and our security has been restored.