[UPDATE (12/9/11): See here for my updated assessment as to US citizens captured abroad.]
On the day that the Senate passed its version of the NDAA, I wrote a post in the morning addressing whether the bill could be read to affirm that detention authority extends to US citizens. Reading the existing language of section 1031 in conjunction with section 1032, I concluded that the best reading of the bill was: yes, section 1031 encompassed citizens. Later that day, Senator Feinstein offered an amendment to the bill in an effort to preclude that outcome, by explicitly altering section 1031 so as to state clearly that citizens are not included. This amendment failed. Still later, she offered a fall-back amendment, altering section 1031 so as to say that it should not be construed as taking a position on the US citizen question one way or the other. That amendment was adopted, and is now part of the Senate bill as the conference on the NDAA gets underway. I later went back and updated my original post to reflect this.
So where precisely does this leave things? Well, right where Senator Feinstein says it does. There are three scenarios in which the government in theory might try to use military detention with respect to a citizen, and the current state of the law is unclear as to two of them.
First, it might try to detain a citizen who is an arms-bearing member of the enemy's forces in a foreign combat zone. Hamdi makes clear that detention authority does extend to that situation already, under the AUMF, and that this is constitutionally permissible (which is no surprise, in my view; In re Territo has long been a standard cite for that same proposition).
Second, the government might wish to detain a citizen found here in the United States, alleging involvement in al Qaeda or another AUMF-covered group. This issue arose with Jose Padilla, an al Qaeda member and U.S. citizen who was arrested on arrival at O'Hare Airport in Chicago and then eventually held for long period in military custody. He challenged that detention through a habeas petition, with mixed results. Suffice to say that the district judge felt that detention authority did not extend to this scenario, that the Fourth Circuit panel hearing his case somewhat avoided the issue by emphasizing the idea that Padilla previously had born arms on the combat zone in Afghanistan and thus was actually similarly-situated to Hamdi, that some observers were confident the Supreme Court would reverse, and that we never found out because Padilla was transferred to civilian custody in order to face prosecution (he was duly convicted and is now in jail). A similar case involving a non-citizen captured in the United States, Ali Salah Kahleh al-Marri, produced a similar result. In short, this is exactly what folks mean when they say that the status quo is unsettled on the question of authority to detain within the U.S.
A third scenario would involve an attempt by the government to hold in military custody a citizen linked to an AUMF-covered group who is captured outside the United States, but not in a hot battlefield context and lacking any prior connection to such combat operations. Say, for example, that Anwar al-Awlaki had been captured in a Special Forces raid in Yemen, rather than killed in a drone strike. We've not had a case like that yet, so it seems to me we'd have to say the law is at least somewhat unsettled as well.
So what does the NDAA have to say about any of this? Nothing at this point, thanks to the Feinstein amendment. For better or worse, the Senate version is explicitly agnostic as to these matters. If it is enacted with that qualification, then the government will be no more and no less able than before to assert detention authority over citizens, and the courts should be no more and no less likely to rule on the matter one way or the other.