The odds of getting reliable information from detainees about the existence of links between the Ba'ath regime and al-Qaeda seem almost nil. Aside from the unlikely event that the detainees would even know anything about such a sensitive and high-level arrangement, it would be difficult to discern whether they are being truthful. And even in that case, such a high-level PIR would almost certainly be classified. If such information were obtained, it would probably have been obtained very early on in the occupation - but that was when we were the least sophisticated in our operations. After that, once the insurgency heated up, I doubt that we were too worried about extracting that type of information. So you've got time, priorities, memories, and skill level all working against the probability that, even if such information were out there, that we managed to obtain it, know whether it was accurate, and made it accessible now by not overclassifying it.

As to whether there were links - either overt or implied - I think the most likely instances occurred in the region where the borders of Kurdish-controlled territory in Iraq met the Iran-Iraq border. There are reports (as to their reliability, I have no idea) about Iran and Iraq not necessarily directly cooperating with Ansar al-Sunna, but generally looking the other way because they had a common enemy in the Kurds. Ansar al-Sunna allegedly had contacts with al-Qaeda and there was allegedly movement of al-Qaeda and Ansar al-Sunna operatives between Afghanistan and Iraq - largely confined to the border region mentioned above. The connection to the Ba'ath regime seems weak, even if these reports are accurate. But they speak to at least some connection. It may turn out that the truth lies somewhere between the two extremes of "no connection" and "slam-dunk evidence of a connection." Now that would really be something.