Chapter V of the Review Group report turns to what we might call the problem of Angela Merkel. It's not about what the legal authorities to spy should look like. It's about what policy structures should govern "Determining What Intelligence Should Be Collected and How." In many ways, it is the portion of the report that will be easiest for President Obama to adopt---and probably easiest as well for some future president to abandon. With a few caveats, most of it seems about right to me.
Recommendation #16 suggests creating "a new process requiring high-level approval of all sensitive intelligence requirements and the methods the Intelligence Community will use to meet them." It suggests "a small staff of policy and intelligence professionals [that] review[s] intelligence collection for sensitive activities on an ongoing basis throughout the year and advise the National Security Council Deputies and Principals when they believe that an unscheduled review by them may be warranted." Recommendation #17 then follows up by suggesting that senior policy makers not use the current static list of requirements that get senior-level review but also examine "any other requirements that they define as sensitive." Recommendation #18 adds that the DNI "should establish a mechanism" to monitor activities to make sure they are consistent with the determinations of these groups---and that this mechanism should include an annual report to be shared with the congressional intelligence committees.
Given the amount of heat the administration has taken for spying on allied foreign leaders, these proposals make sense. An administration ought to know with some specificity what procedures its intelligence community is using to target what sort of foreign friends so that if they ever come to light, the administration is willing to stand by them. The recommendations do not forbid any techniques or requirements; they merely up the level of political accountability associated with deploying them. That seems about right to me.
Recommendation #19 then suggests specific criteria for "decisions to engage in surveillance of foreign leaders":
(1) Is there a need to engage in such surveillance in order to assess significant threats to our national security?
(2) Is the other nation one with whom we share values and interests, with whom we have a cooperative relationship, and whose leaders we should accord a high degree of respect and deference?
(3) Is there a reason to believe that the foreign leader may be being duplicitous in dealing with senior US officials or is attempting to hide information relevant to national security concerns from the US?
(4) Are there other collection means or collection targets that could reliably reveal the needed information?
(5) What would be the negative effects if the leader became aware of the US collection, or if citizens or the relevant nation became so aware?
These criteria seem generally reasonable as well. I have two reservations about them---one involving something they say, the other involving something they do not say. First, limiting the concept of "need" to assessing threats to our national security---as the first criterion does---seems to me unduly narrow. There are good reasons to spy on foreign leaders that have little to do with national security threats: understanding where one stands in bilateral negotiations, for example, or understanding a foreign government's potential receptivity to an as-yet-unadvanced proposal. I would tend to think more broadly in the first criterion about whether there is a need to engage in such surveillance in oder to assess matters with significant importance to our national interest. Second and more importantly, there is a big criterion missing from this list. At least in my view, it matters a lot whether the country in question is one that behaves with restraint in conducting espionage against US leadership. The case for forbearance in espionage is far stronger where staying our hand would honor the idea of reciprocity in relations than where it would operate as a windfall for the hypocrisy of countries who both conduct intelligence and engage in a certain moral preening about being the subjects of intelligence collection. As a general matter, the Review Group too often ignores reciprocity, which is actually a critical aspect of any bilateral intelligence relationship.
Recommendation #20 suggests examining the feasibility "of creating software that would allow the National Security Agency and other intelligence agencies more easily to conduct targeted information acquisition rather that bulk-data collection." The Review Group candidly admits that "[W]e are unable to determine whether this concept is feasible or fantasy" but suggests that "It might reduce budgetary costs and political risk if technical collection agencies could make use of artificial intelligence software that could be launched onto networks and would be able to determine in real time what precise information packets should be collected." I don't know either whether this is feasible or fantasy and agree that it would be useful to know. So sure. Examine away.
Finally, Recommendation #21 suggests exploring "understandings or arrangements" with other governments "regarding intelligence collection guidelines and practices with respect to each others' citizens (including, if and where appropriate, intentions, strictures, or limitations with respect to collections)." The idea here is that the closer the intelligence partnership, the more reciprocal restraint in intelligence collection is possible---the Five Eyes arrangement being the most extreme example. I am frankly pessimistic about the number of opportunities for additional such arrangements, but I agree that they are always worth exploring with appropriate governments. And I agree with the Review Group as well that the posture of the US intelligence community should be one of openness to "explor[ing] the possibility reaching similar arrangements and understandings with a small number of other closely allied governments," even though "very few new such relationships are likely in the short to medium term." Indeed, my reservation about Recommendation #19 is that it does not adequately incorporate the idea at the core of this recommendation---which is that intelligence relationships are negotiated, not matters of unilaterally imposed restraint in the face of significant national interest.
In short, Chapter V contains, in essence, no new proposals for legal changes---just a series of suggestions that the US should be more wary of the political consequences of collection and should create executive policy oversight structures to head off those consequences. In light of the post-Snowden disaster, this basic approach seems wise, and the group's suggestions seem generally on target.