In the current wobbly post-Cold War phase of American global engagement, it seems that mistakes tend to be committed more out in the open, or they go public sooner. This presents planning and political problems that our leaders have only begun to sort through.

Take the pair of mistakes being attributed to the CIA: the bombing of a building in Sudan that was thought to be a chemical war facility but may have been just a pharmaceutical factory, and the targeting of a complex in Belgrade that was thought to be a Yugoslav army supply and procurement headquarters but turned out to be unquestionably the Chinese Embassy.

The Sudan bombing last August, besides apparently failing to accomplish its mission of striking at terrorist Osama bin Laden's chemical works, subjected the CIA to terrific embarrassment for its method of choosing targets. Part of the embarrassment is the successful suit that the factory owner, a Saudi, subsequently filed against the United States for his $30 million loss.

In Belgrade in May, the Chinese decided to make a big political storm out of the American bombing of their embassy. They worked from the allegation -- for which they offer no more than thin circumstantial evidence -- that the United States deliberately targeted the embassy.

Frankly, I find it a tale out of Oliver Stone. But there is no denying that the CIA's apparent Sudan errors invite the conclusion that an agency that could get it wrong once could get it wrong twice. It's a sad fate when you consider that, according to the director of central intelligence, George Tenet, the Yugoslav arms agency was the single target selected by the CIA out of the 900 struck by NATO and the U.S. military's European Command during the 78-day Kosovo bombardment.

Obviously, the CIA's targeting people need help. It's hard to think of a single government agency that has tripped up in so many ways central to its designated national purpose. Keep in mind, too, that the agency and the larger "intelligence community" that it heads may have encountered some further embarrassments or disasters that are not yet part of the public record. Congressional investigations of the Sudan affair are still going on.

A dismal irony is at work here. In the past 10 years our intelligence agencies have been constantly urged to set aside the old Cold War priorities directed at a Soviet threat and to turn to post-Cold War priorities such as terrorism and ethnic strife. Well, the Sudan bombing was conducted in the service of the struggle against terrorism and the Belgrade bombing to sort out the disintegration of Yugoslavia. Our people had their eye on the ball, but, in these instances anyway, they dropped it.

Have they now picked it up -- drawn the right lessons? Intelligence chief Tenet, taking "ultimate responsibility" for the accidental bombing of the Chinese Embassy, defined the problem as "targeting in urban areas." He identified the "critical lesson," murkily, as providing "an accurate appreciation of our confidence in the location of a target and [of] the evidentiary basis for how that location was determined."

It is not clear to me what "ultimate responsibility" means if it does not mean that the person responsible for an admittedly "major" error, which killed people and had severe political consequences, intends to resign. In fact, what Tenet is taking onto himself is not ultimate responsibility but limited responsibility. Or perhaps procedural responsibility, since his remedies run to technical procedures and not to the questioning of bombing, as he puts it, "in urban areas" -- where other buildings are apt to clutter the landscape, not to speak of people, who, after all, tend to live "in urban areas."

Nor, in his testimony to Congress, did Tenet address the question of bombing in urban areas that happen to be in someone else's country -- bombing in a sovereign state, for instance, a state with which we are not at war. I think a strong case can be made for such violations of sovereignty in order to serve the interests of American citizens. But it is not a self-evident and unanswerable case, and it needs to be made, seriously and respectfully, not so much by an agency chief as by our ultimate political authority, the president.

Here lies the political context of post-Cold War security policy. It must be conducted on the premise of an open society with meaningful accountability and full scrutiny of means as well as ends. A tough test for tough times.