Intelligence Incompetence and Probabilistic Terrorism

When news of Hamas' 10/7 attacks broke, one of the first things I wrote was straightforward: "a massive intelligence and operational failure the likes of which Israel's security services have never seen in my lifetime." Now, the New York Times has a blockbuster report detailing exactly the magnitude of these failures, including the fact that Israeli intelligence did have knowledge that such an attack was being planned but nonetheless dismissed it as infeasible and unlikely to actually occur.

As always, I bow to nobody in my utter contempt for the Netanyahu administration. And there's little question that no matter what curve you grade them on, this was a spectacular failure of Israel's intelligence and security apparatus. Nonetheless, even in these cases one can't claim too much benefit from hindsight. And the "lessons" that many are implicitly seeking to draw from these reports are ones that I worry are neither reasonable ones nor, ultimately, salutary ones.

Every potential event, until it actually happens, is a probability cluster. It's something that might happen, but also, might not. Proximity and evidence can make the probability more or less certain; and probability will also inevitably be contingent on the presence or absence of certain interventions (whose effects, again, can only be measured probabilistically). Most importantly, a prospective future event that is successfully forestalled by definition never happens, and so we can never know whether its probability cluster would have borne fruit. In these deliberations, we're always weighing the reality that did occur against a hypothesized alternative that never did, and so it's understandable and inevitable that the former is going to carry disproportionate weight.

Consider the oft-fantasized prospect of assassinating Hitler in 1933. Let's stipulate that doing so would have averted the worst horrors of the following decade: no World War II, no gas chambers, no Holocaust. Knowing what we know now, it's essentially indisputable that this would have been a justified and salutary act that could have forestalled incalculable levels of human misery. But the problem is that, if Hitler was assassinated in 1933, they wouldn't -- they could never -- "know what we know now." They would never know the horrors that would have been averted. All they would know is that the new leader of Germany was assassinated, as well as whatever fallout resulted from that act. Without knowledge of what happened in the prime timeline, it'd be virtually impossible to persuade anyone that killing Hitler meant averting the most destructive war and the most brutal genocide of the modern era. All of that would simply be a probability cluster that now wouldn't ever come to fruition. Indeed, imagine, as I think is plausible, that the assassination of Hitler resulted in a series of Kristallnacht style riots against Germany's Jewish population, but that things stopped there -- no World War II, no gas chambers, no Holocaust. The history books in our alternative-universe probably would treat the assassination of Hitler as a disaster for Germany's Jewish community. They'd never know differently.

And the problem is worse than that. I have a relative who periodically sends me emails regarding the latest emerging global leader who's spouted off some antisemitic nonsense. Typically, I do my best to assuage her that this person is fringe, is not actually that influential, and is not in a position to do anything especially dangerous. She then inevitably replies to me by saying "that's what they said about Hitler." And the thing is -- she's right! That is what they said about Hitler! The problem is that that's also what they said about a lot of people who didn't turn out to be Hitler. If we violently removed from power every person who possibly could become Hitler, that's a recipe for geopolitical chaos -- and most of the violence would be directed in cases where it wasn't necessary, where the threat would dissipate on its own. In any event, we'd never know which case was real and which ones were false alarms.

Return to the matter at a hand. Roll back the clock to October 1. Israel has intelligence suggesting Hamas is planning an attack like Al-Aqsa Flood. In our parallel universe, they do take this intelligence seriously. What results? Most likely, some sort of preemptive attack or assault into Gaza meant to degrade Hamas' military capacity so that the attack cannot be launched. Would that have been justified? With the benefit of hindsight, we know that such an operation would have forestalled a much greater evil, both in terms of stopping the Hamas attack but also almost certainly avoiding the blistering Israeli response that's occurred over the past two months. But that's the problem -- if the operation is successful, none of that future would ever come to past, so the only thing people would see would be the preemptive, "unprovoked" Israeli attack into Gaza, justified at most by a hypothetical probability that if they hadn't acted, Hamas would have launched a brutal terrorist attack of its own.

There's no getting around this. Perhaps, you say, Israel could release the intelligence showcasing Hamas' plans. But again, try to rewind your mind to October 1. How likely is it that Israel would be believed? Many would dismiss the intelligence as warmongering propaganda. There's rampant denialism about the atrocities Hamas actually did commit; can you imagine how likely people would be to accept the proposed destruction of a hypothesized Hamas attack that never occurred? And even persons who did not believe that Israel was flatly lying might be quite likely to be dismissive of Hamas' actual capacity to launch the attack -- sure, they might fantasize about it, the argument would go, but it's rank fear-mongering to act as if they could actually pull it off against the Most Powerful Military in the Middle East. I've seen more than a few commentators suggest that it was "racism" on the part of the IDF to assume that Hamas was not operationally capable of successfully conducting Al-Aqsa Flood; but if we had publicly pitched this scenario on October 1, realistically it's Palestine sympathizers who would have been most dismissive of the prospect. The right-wing hawks, whatever their other sins, are not typically known for understating Hamas' threat; it's the left that has long treated Hamas' military capacity as essentially a null entity against the Israeli juggernaut, its rockets as glorified sparklers and its calls for antisemitic annihilation as slightly-overwrought rhetoric from impassioned revolutionaries. The uncertain probability of a future Hamas attack -- one that, in our alternative universe, never would happen (because it was successfully forestalled) -- will have little persuasive power against the reality of the Israeli intervention.

I don't want to suggest, however, that this is merely a PR problem. Again, the troubles run deeper than that. I have to assume that Israeli intelligence gets reports on the daily of incipient Palestinian militant activity, including desires for large-scale, major operations. Most, it's almost certainly the case, fizzle out without ever reaching fruition; just as most potential-Hitlers never actually become Hitler. For that reason, it is both good and necessary that Israel not respond to every one of these incipient threats as if it is the next October 7, for the same reason why it's good and necessary that we not preemptively assassinate every global politician who "could be Hitler". An Israel which is single-mindedly determined to stop the next 10/7 would be beyond aggressive, it would be lashing into the Palestinian territories with overwhelming force every week -- way beyond even the baseline levels of military activity and suppression that occur as part of the occupation, and in most cases it'd be in response to intelligence about activity that would have never led to a 10/7 anyway. Such behavior, in addition to its catastrophic impact on the Palestinian population, would have unpredictable (or perhaps all too predictable) knock on effects on Israeli security -- it's not sustainable, which is a good thing, because if it was sustainable it'd be horrifyingly dystopian.

Again, I'm not trying to downplay the magnitude of the intelligence failure in this case. The inevitable uncertainties and probabilistic reasoning I identify above is an inherent feature of life; one has to make better or worse choices within that uncertainty, and the Israeli security apparatus made a terrible choice. That assessment is not, I think, merely a product of hindsight. But recognizing that Israeli intelligence could have made a better choice knowing what they did at the time doesn't mean the problem of hindsight isn't real. A "good" choice by Israel likely would have been met with harsh condemnation, because (thankfully) people would never know what hell had been averted. And even that stipulation is too generous, since in the real world where we don't know and never will know what events we've averted, we can't actually know if we're killing the metaphorical Hitler or just unnecessarily engaging in superfluous preemptive violence.