Majid Khan is a CIA detainee. In meetings with his attorneys, Khan has apparently made allegations concerning his treatment during his detention. Yet such information, even as detailed by Khan himself, is presumptively classified, and his attorneys are apparently barred from relating Khan's claims to Congress. Marty Lederman finds this "absurd," and I am inclined to agree. As Lederman asks, "Even if the classification [of the techniques] were itself valid, can it really be the case that the persons against whom the CIA employed its methods may be prevented from disclosing such historical facts to the public?"
I can see an argument for barring Khan's attorneys from disclosing information that might reveal classified operational details (e.g. how Khan was identified and captured) or sensitive intelligence information (e.g. what Khan told the CIA or learned the CIA knows). The disclosure of such information might compromise national security. Yet at issue here is nothing more than information about how Khan has been treated by the U.S. government — information Khan would almost certainly be able to reveal were he not currently detained and which might help inform the public debate about how detainees are and should be treated. (I recognize that Khan could lie about his treatment, but this risk would hardly seem to justify the prohibition at issue; if he were to lie, the government could dispute his claims.)
The Khan case has disturbing similarities with the Higazy affair. As regular readings of Howard Bashman's How Appealing know, Abdallah Higazy is an Egyptian national who was detained by the FBI and who, in the course of interrogation, made a false confession. (Apparently the FBI's interrogation techniques "worked.") Among other things, Higazy claims that the FBI threatened to identify Higazy's family to Egyptian authorities. Fearing his family could be tortured, Higazy alleges this threat induced him to confess. After his release, Higazy sued and, in an opinion last month, the U.S. Court of Appeals for the Second Circuit held that his suit could proceed.
Interestingly enough, the official public version of the Second Circuit's opinion is redacted. Although there is no dispute that Higazy was innocent of any wrongdoing — he's an engineering student, not a terrorist — some of the information concerning Higazy's FBI interrogation was filed under seal. Some of this information was included in the original version of the Second Circuit's opinion, which (as Bashman recounts here) was initially posted on the Second Circuit's website (and later posted on How Appealing). The opinion was later withdrawn in favor of the redacted version, and a court official asked Bashman to remove the opinion from his site, a request Bashman refused.
As in the Khan case, the "secret" material concerns the U.S. government's alleged (mis)treatment of a detainee, not the sort of information for which classification could be justified on national security grounds. (See Patterico's analysis here.) The disclosure of such information could be embarrassing to the U.S. government, to be sure, but that would hardly justify keeping such material classified or barring public disclosure. To the contrary, public policy often explicitly encourages the disclosure of such information by protecting (and sometimes even rewarding) whistleblowers and others who disclose potential government wrongdoing.
I recognize that there is much information that, at least for the time being, needs to be kept secret for national security purposes. There may even be an argument for concealing some details about highly sensitive interrogations (though I am skeptical). Yet I see no reason why the specific allegations about how Khan and Higazy were treated should be kept from the public. There are many reasons to classify information and enforce government secrecy, but fear of embarrassing the government with allegations of misconduct is not one of them.
UPDATE: Several commenters argue that there is a valid basis for keeping the details of interrogation methods secret. I agree with this, though I also believe that there are countervailing interests against which such concerns need to be balanced. There are reasons why we would not want our most feared enemies to know all of our interrogation protocols. At the same time, political accountability requires some amount of disclosure (particularly to Congress).
Note, however, that this is not what is at issue in either the Khan or Higazy incident. In these cases, what is at issue is the disclosure of what actually occurred during interrogation. Set aside the allegations of illegal conduct by the interrogators (though this is clearly relevant), in neither case would disclosure give terrorist operatives a detailed understanding of our interrogation protocols. It is simply implausible that the Khan interrogations, for instance, exhausted the universe of potential authorized approaches, and future detainees would have no assurance that they should only expect (and prepare for) what was done to Khan. For this reason, I do not think that the argument for classifying interrogation techniques establishes the need to keep Khan's (or Higazy's) claims secret.