Seven security experts spoke at the Apollo Theatre on Thursday on a panel titled Navigating National Security: Intelligence Gathering in the Digital Age. Five of the panelists, Jennifer Sims, OC ’75, Robert Jervis, OC ’62, Judith Klavans, OC ’68, Diana Wueger, OC ’06, and Joe Alhadeff, OC ’81, spoke with the Review about the PATRIOT Act, data and Edward Snowden.

Do you think the PATRIOT Act should be extended, modified or allowed to exist?

Jennifer Sims: I think it ought to be extended. Not permanently. I do think it ought to be extended, and it’s coming up, as you know, in June. I have concerns about aspects of it. So the three provisions that are sunsetting: one of them is Section 215, everyone knows about that, the other is roving wire taps, that’s coming up, and the other is the “Lone Wolf ” provision.

My greatest concern about Section 215 has always been the gag rule. If I had my way, there would be no gag rule. I believe that is a step too far because I think forcing people not to speak about what the government has asked them to do and give up, even when they think it is deeply wrong, is undemocratic. Now, having said that, there’s a counter-intelligence cost. There’s a security cost to it, and I’m very cognizant of that because if people can speak about what the FBI is asking of them, for example, then the adversary knows they’re being chased, so you complicate the counter-intelligence problem, but I think that there ought to be mechanisms. A gag rule is a very, very serious thing in a democracy.

Is there a national security risk to collecting too much data and, as a result, losing the information that’s actually important in all the noise?

Judith Klavans: I like big data. I think that all information, bring it on, and the reason for that is statistical. If you can get a statistical basis for what is ordinary, then you know what is extraordinary. So you know it’s outside of what you normally expect. So just taking the case of language. If we talked about your profile as a writer, we each have language that we gener ally use. If I’m trying to pretend that I’m you, you’re going to know it in a heartbeat. If you have lots of his language and lots of my language, you’re going to really look different. Joe uses words I would just not use. I just don’t use them. They’re kind of legally words. That’s why I’m sitting here saying, ‘How am I interpreting this?’ — because we have different kinds of vocabulary uses. So the more that we have as baselines for people and profiling people, it’s almost like a fingerprint. You can tell who you are by how you write. One of the examples Jordan and I were just talking about on the way over is that, for cybersecurity for example, if you have an example of a computer code that a hacker or cybercriminal has written, you can look for code just like you can look for an essay that you write and I write that looks like that and guess that it might be the same author and guess that it’s a cybercriminal. And then look at where it’s coming from, what its intention is and what it’s trying to do. So I do not think we can collect too much information. How we protect it is another issue.

JS: I’d add one other thing though, and that is one of our vulnerabilities in the U.S. intelligence system is what I call the all-source analysis vulnerability. We have this tendency to always think that all-source analysis is best. So if you get a piece of intelligence, and it seems really important, send it back to the National Counterterrorism Center or CIA, it gets put into [a] big database and the analysts then look at the database and the little piece nugget then gets lost in the pile of nuggets where people are trying to connect dots, whereas if there was, at the front end, people making a discrete choice “Is this actionable intelligence today? Do I know where to send this today?” — [which] is so significant that a decision maker or a warrior or other person needs it immediately. If we’re going to stop something bad from happening then, yes, you may send a copy back to [a] central database, but you could also hand it off to someone who could use it right away. Now … there’s a danger with that because it isn’t necessarily all source, and it might be wrong. But that’s a risk.

Joe Alhadeff: Part of that also goes to so that if one of the reasons that you’re collecting a mass amount of information is that it’s because you can’t pattern the norm, you don’t know what normal looks like. You don’t know what normal looks like, you can’t find an outlier. But part of the question is, “Do you need all of that information in its most identifiable form to pattern the norm?” Because in theory you could use some levels of aggregation or some de-identification and still have very useful information because it could give you the granularity you need, but that again tips your hand to the people that are using this “Well I don’t need to know all of this.” All of the sudden people know what you kind of know and what you don’t know and so the ability to feed you misinformation inside of that leads you to greater danger. But as you can tell, this isn’t a black and white.

Yes or no: If you were President Obama, would you pardon Snowden?

JS: Pardon Snowden now or later after he’s convicted? [Laughs.] No.

Robert Jervis: No.

JK: No.

Wueger: No.

JA: No.

RJ: Even if one approves of what he did, which is a different question.

JS: I do not approve of what he did. I do not because I think he undermines our oversight institutions in a horrible way. We have mechanisms for whistleblowers to use who can protest within our system. To go outside the system, to deny the institutions the ability to work is to undermine our democracy in a very fundamental way. I protested Watergate; I protested the Vietnam War; I marched on Washington but it was always to improve our institutions, to reassert the importance of our democracy and our political institutions and our representatives in Congress and Snowden has not appeared before our representatives in Congress. He has not submitted himself to the democratic processes we’ve put in place for just this kind of thing.

He sowed so much distrust in the American political system about the intelligence community, and it’s this amorphous distrust. It’s hard to have a reasonable debate. I agree with some of the Snowden view of overreach by the intelligence community. What he did was unforgivable. I’ve worked too long to try and work within these institutions to make them responsible and accountable. And for him to attack that oversight system from outside it is reprehensible to me.