Artificial Intelligence

I have one question for National Intelligence Director James Clapper and his predecessors. As we all know, the infamous 2007 National Intelligence Estimate asserted with “high confidence” that Iran had halted its work on nuclear weapons development, and no subsequent NIE ever reversed that judgment. Yet fast forward seven years, and the latest annual intelligence assessment asserts that “Tehran has made technical progress in a number of areas … from which it could draw if it decided to build missile-deliverable nuclear weapons,” and consequently, it now “has the scientific, technical, and industrial capacity to eventually produce nuclear weapons” should it so choose. So here’s my question: If Iran stopped its weapons development effort seven years ago, how did it happen that since then, it has made precisely the kind of technical progress that now enables it to build a nuclear warhead whenever it chooses?

There are two plausible answers to this question. One, the work never really stopped, or resumed at some point in the last few years, and U.S. intelligence agencies simply missed it. That’s certainly possible; intelligence agencies aren’t omniscient, and it’s unrealistic to think they will never make mistakes. A more troubling possibility is that since intelligence rarely reaches the level of absolute certainty, the available information was misinterpreted due to political bias–a desire to avoid military action against Iran, and hence to avoid interpreting Iran’s behavior in a way that might necessitate such action.

But the answer offered by the Obama administration strains credulity: that Iran really did stop its weapons program and never resumed it, but somehow, mysteriously, nevertheless made major technical progress over the last seven years of precisely the kind that now enables it to build a nuclear warhead anytime it pleases. Even a two-year-old wouldn’t buy that.

The real problem, however, isn’t what this says about the past, but what it says about the future. After all, for years, opponents of attacking Iran’s nuclear program have argued that Tehran hasn’t yet decided to make a nuclear weapon, and if it ever does, the U.S. will know in enough time to stop it before it succeeds. Therefore, there’s no reason for either America or Israel to take military action now. Yet how can either Americans or Israelis have confidence that U.S. intelligence will detect a nuclear breakout in time if, for the past seven years, it has either missed all the signs that Iran was continuing to make “technical progress” toward weaponization, or deliberately ignored them out of a desire to avert military action–a desire that, judging by both words and deeds, remains the administration’s top priority?

The answer, of course, is that they can’t. And the lesson for Israel is clear: It cannot rely on U.S. promises to stop Iran from getting nukes, because these promises are based on the faulty assumption that U.S. intelligence will uncover a “smoking gun”–the kind of irrefutable proof that can’t be argued away–in enough time to take action. Hence the day is coming closer when Israel will have to make a fatal decision: attack Iran itself, or learn to live with a nuclear Iran.