In mid-2009, just after Neda Soltan was shot down on a Tehran street, I was working with a group of hackers attempting to provide covert communications support to the Iranian dissident movement. I blogged about it at the time, received my second death threat as a result, and had a couple of interesting conversations with the FBI. Eventually I stopped working with the group; they made what I considered some serious mistakes of direction, and I was anyway beginning to doubt the principal’s claims of having an extensive contact network on the ground in Iran.
The group I was working with was deploying stealthed HTTP proxies and Tor nodes; my main technical contribution was a Squid configuration tuned to the purpose in ways I won’t discuss because the techniques might be useful against another tyranny someday. We knew of other groups using different technology; one seemed to be organized around a program called Haystack and its designer, a guy named Austin Heap.
I had brief contact with Heap during that period. Now I learn that he may have been a fraud. The Great Internet Freedom Fraud all but accuses him of this.
I’m blogging about this because, on the one hand, I had a few suspicions about Heap at the time. On the other, neither my interactions with him nor (in my judgment) the public evidence quite supports the theory that he set out to scam people. For whatever it’s worth, I think he started with good intentions, talked a bigger game than he was really capable of executing, and got in over his head.
There was a lot of that going around – including, very possibly, in the group I was working with. To this day I don’t know what, if any, impact we actually had. Could have been a lot. Could have been nothing. All I know is that someone apparently affiliated with the Iranian regime threatened to have me killed over it, and the FBI believed the threat was credible enough to talk to me about the whole deal. But that threat, too, could have been a phantom or fraud, and well I know it.
Why am I blogging about this now? Partly because I think the Slate article overreached a bit in its implied accusations against Heap. I was in the trenches with him and others who were trying to help the dissidents, and I think he’s owed just a bit more benefit of the doubt than he’s getting there.
But mainly I want to point out what this episode teaches about the epistemic problems of operating in a stealth mode, among a group that is trying to do operational security. These problems are shared by spies, terrorists, and groups like the ones I was working with who are gaming against spies and terrorists.
The problem is this: to protect your network, and yourself, you have to accept that you are going to have relatively little information about what your network partners are doing and what their capabilities are. In this instance, I didn’t want to know details about the scope or nature of our on-the-ground network in Iran; I was, for some crucial weeks after the Neda Soltan shooting, our public contact person, and thus the most exposed to pressure by our adversaries.
The obvious problem was, my rationally-chosen ignorance left me unable to form judgments about whether people in my network were lying to me. More subtly (and here we get back to what I think happened with Austin Heap) it left me unable to form judgments about whether they were lying to themselves.
Normally, I think I’m pretty good at detecting self-deception – face to face. There are tells in affect, tone, and body language that you learn to spot after a few decades of peoplewatching that tell you a speaker is actively suppressing doubt about his own utterences (and these are slightly different than the tells for outright lying). I’ve actually had to practice this; it’s one of the most important person-to-person leadership skills. But it’s much more difficult to spot self-bullshitting over text links and chat, and I have little confidence that I can do that reliably.
Thus, my ties to this network depended on a high level of faith in my contacts. It couldn’t be otherwise, because trying to gather information that might have allowed me to reason out the odds better might have jeopardized them, jeopardized the dissidents I was trying to help, jeopardized me, and blown the network. As a direct consequence, my participation was fragile; when my contacts said and did things that eroded my faith in their judgment, I dropped out.
Here’s the next turn of the screw: I think living in that kind of informationally poor shadow-world makes it more likely that a well-intentioned but only marginally competent person will overcommit to others and himself in order to increase his influence, and eventually con himself into a role he can’t sustain. Because who has the information to call bullshit soon enough to stop the escalation before it gets past little white lies?
I suspect very strongly that this is the trap Austin Heap fell into. And I’m not sure my contacts weren’t down the same hole and my effort wasted. I don’t think I can assign less than a 30% probability to that outcome.
Still – even with that uncertainty I’d do it again, death threat and all. Because I’m a man and an American and a libertarian, and all three of these entail duties. In the course of those duties, there are some sorts of risks that must be run; honor and ethics require it. The risks I took, and am still taking as a result, were among them.
I don’t think my experience was unique; people who fight for good causes in shadow have, no doubt, been experiencing these problems since history began. But I’m not sure I can extract any larger lesson here. I don’t mean to excuse whatever lies Austin Heap may have told, but I do mean to suggest he may well have been his own first victim.