Now you see it. Now you don’t.
This is not a magician’s incantation. It is a description of retroactive classification, a little-known provision of U.S. national security law that allows the government to declassify a document, release it to the public, and then declare it classified later on. Retroactive classification means the government could hand you a document today and prosecute you tomorrow for not giving it back. Retroactive classification can even reach documents that are available in public libraries, on the Internet, or elsewhere in the public domain.
The executive branch has used retroactive classification to startling effect. The Department of Justice, for example, declassified and released a report on National Security Agency (NSA) wiretapping only to declare, years later, that the report was once again classified. The journalist who had received the report was threatened with prosecution if he did not return it. Retroactive classification has also targeted government documents revealing corruption in Iraq, violence in Afghanistan, and mismanagement of the national missile defense program. In each of these cases, the government released a document in an unclassified form through official channels—not through a leak—and then turned around to classify it.
This practice would be troubling enough if it actually removed the document from the public domain. But in the Internet Age, once a document is released to the public, it is often impossible for the government to retrieve it. While retroactive classification does not remove the document from the public domain, where our enemies can access it, retroactive classification does remove the document from the public discourse, prohibiting members of Congress, government auditors, and law-abiding members of the public from openly discussing it.
In the ongoing debate about the balance between secrecy and transparency in government affairs, retroactive classification tests the limits of the government’s ability to control information in the public domain. The questions raised by retroactive classification go far beyond those raised by the WikiLeaks and Edward Snowden disclosures. In those cases, the information remained classified even though it was widely available in the public domain. A similar situation occurs with retroactive classification when information in the public domain becomes classified. The difference is that in retroactive classification, the government initially released this information in a non-classified form and only later decided to classify it. This difference makes retroactive classification much more complicated from a legal standpoint because it involves the government’s going back on its initial classification decision. Retroactive classification thus forces us to ask what limits, if any, exist on the government’s authority to control information. Can the government reach into the public domain to make a secret out of something it has already disclosed? Are we obligated to go along with retroactive classification decisions? What are the implications beyond national security law? This Article takes up these pressing questions.