The litigation over former National Security Advisor John Bolton’s memoir has taken a long and winding road. It began with the Department of Justice filing a breach of contract complaint against Bolton last Tuesday, like the government does in every other prepublication review case: the first step of a lawsuit designed to ultimately seize any money he makes from allegedly breaching his non-disclosure agreements (NDA). The case took an unexpected turn when the DOJ filed a motion for a temporary restraining order (TRO) on Wednesday night seeking to stop the book from being sold. It then veered again when Bolton filed a motion to dismiss the entire case late Thursday night. On Friday, Judge Royce Lamberth held a hearing on the government’s TRO motion, and on Saturday morning he denied the motion. In significantly less time than most federal cases take to properly serve the defendant, this case has already seen three motions, three amicus curiae briefs, 10 sworn declarations (of which two are classified), an amended complaint, a public hearing, an ex parte hearing, and an opinion.
While Bolton’s book is now being sold freely, this case is far from over. Unlike some of my colleagues, I believe that the government’s breach of contract case is pretty solid. When you remove the flurry of activity around the government’s ill-advised attempt at prior restraint, it becomes a textbook prepublication review case, and those almost always result in a resounding government win. Bolton still faces the likely prospect of losing any proceeds he might earn from the book, including future proceeds from speaking engagements and the like.
Bolton tries to avoid this result in a few ways, but, while some of them are admittedly creative, I don’t believe any of them will win the day. Over the course of the week, Bolton tried out a few arguments for why he did not breach his contract. First, he claimed that Ellen Knight—the National Security Council (NSC) official initially performing the prepublication review—told him there was no more classified information in the manuscript, leading him to believe that his duty was complete. Second, he claimed that Michael Ellis—the senior NSC official who re-reviewed the manuscript after Knight was done—did not have the authority to do so. Third, he claimed that the NSC was only allowed to redact information which was classified or in the process of being classified at the time he submitted it for review. Fourth, he claimed that he was not even required to submit his manuscript at all because he did not believe it contained classified information.
I will address these arguments in reverse order.
Did Bolton Have to Submit His Allegedly Unclassified Manuscript?
Bolton’s last argument is the most easily dispensed with. Bolton argues that his NDA only required him to submit classified information for prepublication review, and that since he deliberately avoided including classified information, he was not required to submit it for prepublication review. This is a ludicrous argument. The entire prepublication regime is based on the fact that under the current law, an author is simply not allowed to make a call on whether his book contains classified information. Authors are required to submit all works for prepublication review to allow other officials to decide if it contains classified information. There are a few requests for this. For example, another official might know that a particular sentence reveals a classified fact the author was unaware of. Alternatively, another official might realize that a statement should be classified even though it is currently unclassified.
The most obvious problem with this argument, however, is that it ignores one of the key messages of prepublication review. Intelligence employees are constantly warned that they should never knowingly include classified information in a manuscript, especially if they are not typing it on a secure computer. According to prepublication review officials, the purpose of prepublication review is to remove any classified information an author accidentally included. It is not a mechanism for getting information declassified. As a veteran national security official, Bolton would be well aware of these admonitions, which means that he could not have been under the impression that he only had to submit classified manuscripts for prepublication review. If you are not allowed to knowingly write a classified document for prepublication review, you cannot knowingly submit a classified document for prepublication review. If you cannot knowingly submit a classified document for prepublication review, you cannot reasonably believe that you are only required to submit classified documents for prepublication review.
Can the Government Retroactively Classify Information During Prepublication Review?
Bolton argues that the plain language of his NDA states that the government can only redact information that was either classified or in the process of being classified at the time it was submitted for prepublication review. This is not the case.
As I discussed above, it is not uncommon for a prepublication review official—who is an Original Classification Authority in his own right—to classify information in a manuscript which was previously unclassified. Retroactive classification is a fact of life in the national security world. Sometimes phrasing something a particular way inadvertently reveals a piece of classified information. Sometimes the world changes, and things that were not dangerous before now are. Sometimes the first person to see the information simply made a mistake in not classifying it.
And sometimes the prepublication review official is the first person to ever see the information. The rules explicitly specify that even works of fiction must be submitted for prepublication review. If a work of fiction—which previously only existed in the mind of the author—can be classified, then by definition it can only be retroactively classified after the document was submitted. Therefore, Bolton’s argument cannot be correct.
For example, imagine that a Central Intelligence Agency (CIA) analyst writes a thriller novel in which the MacGuffin is a drone which can track a target’s pheromones. This device is a total invention of the author’s imagination. However, unbeknownst to the author, In-Q-Tel has invested in just such a device. During prepublication review, the reference to the fictional device would be redacted as classified because a reader might assume that the author—a known CIA analyst—was describing a real device. Alternatively, imagine that such a drone does not exist but CIA does not want hostile forces to believe that one does because it would set off an arms race, thereby harming national security. The completely fictional device would be redacted as classified, whether it existed in real life or not.
As a side note, many commentators have claimed that the fact that the government has said that the information is classified means that Bolton’s allegations must be true despite President Donald Trump’s protestations to the contrary. This, too, is not the case, for the same reason. If national security could be harmed by a foreign power believing that a false statement in Bolton’s book is true, that statement could be classified. There is no requirement that only true information may be classified. For example, one of the reasons that the photos of Osama bin Laden’s body and burial were classified was not because they themselves contained any classified information, but because a hostile actor could Photoshop them to make it look like his body was desecrated. According to this logic, such an altered photo—which would be demonstrably false—could be used to harm U.S. national security and would therefore be classified. Therefore, the unaltered photo must also be classified. This train of thought only makes sense if you accept that a faked photo could be classified because of the harm its release could cause to national security if someone thought it was real. By the same token, a brazen lie in Bolton’s book could still be classified if national security would be harmed by someone believing it because of who he is.
Did Michael Ellis Have the Authority to Re-Review the Manuscript?
Of all of Bolton’s arguments, this exceedingly narrow one is the most meritorious. Ellis became the NSC Senior Director for Intelligence on March 1, 2020, and became an Original Classification Authority (OCA) that same day. On May 2, Ellis began his re-review of Bolton’s manuscript, which he completed on June 9. However, he did not complete the annual security training required by Section 1.3(d) of Executive Order 13,526 until June 10.
It is generally understood that prepublication review must be conducted by an OCA. Bolton’s argument relies on one of the Executive Order’s implementing regulations, 32 C.F.R. § 2001.70(d)(2), which states, “Original classification authorities shall receive training in proper classification and declassification prior to originally classifying information and at least once each calendar year thereafter.” According to Bolton, this means that Ellis could not originally classify information before he received his first training, which would mean that he lacked the authority to conduct the re-review.
This is not a frivolous argument, but I believe it will not ultimately prevail. The question of whether or not these implementing regulations apply to the White House has been hotly contested for over a decade, most visibly when former vice President Dick Cheney asserted that the Information Security Oversight Office had no authority over his office. This remains an unsettled area of the law, and until now it was one which was unlikely to be decided by a court, since one part of the executive branch cannot typically sue another part. If Judge Lamberth decides that this distinction is relevant, however, we could receive some clarity.
However, courts employ a doctrine known as constitutional avoidance, which basically means that they will try their best to resolve a case before reaching thorny constitutional questions like respective authorities within a co-equal branch. In this particular case, Judge Lamberth does not need to reach this question, because the government claims, “After completing the training, Ellis reviewed his work and concluded that the information he received in the training did not alter his decisions.” In other words, even if he lacked the authority to perform the review on June 9, he had the authority to perform it on June 11. (However, this presumes that he reached this conclusion before this litigation was filed, and if that is not the case, it becomes a closer question.)
Was Bolton Allowed to Rely on Ellen Knight’s Words?
Bolton’s key argument relies on the assertion that Knight—the NSC official conducting the prepublication review—informed him that the manuscript no longer contained classified information. Bolton testified under oath that she informed him on April 27, “That’s the last edit I really have to provide for you.” He understood this to mean that the review was complete and that all that remained was “the customary pro-forma letter confirming that the book contained no classified information and was cleared for publication,” which he testified she said “might be ready that afternoon.” This version of events is supported by the government’s complaint, which states, “On or around April 27, 2020, Ms. Knight had completed her review and was of the judgment that the manuscript draft did not contain classified information. Ms. Knight informed NSC Legal of the status of the review.”
According to Bolton, even though he did not receive the “customary pro-forma letter,” he was free to publish because she had informed him that the book contained no classified information. Furthermore, as his argument goes, he was free to ignore the NSC’s efforts to re-review the book because he did not have a reasonable belief that the book contained classified information, again based on Knight’s alleged statement.
While attractive on its face, this argument does not hold up in light of the practical realities of prepublication review. Review is not complete until everyone says it is complete. For instance, in one of the landmark prepublication review cases of this century, the Army Reserve—which controlled the author’s clearance and therefore conducted the prepublication review—officially cleared Lt. Col. Anthony Shaffer’s book Operation Dark Heart for publication, but before the book was printed, the Defense Intelligence Agency (DIA) stepped in and retracted the approval, claiming that it had not yet had a chance to conduct a review. This was legal, and as a result, Shaffer ended up submitting the book to the DIA and ultimately suing the agency over the matter, and the Defense Department purchased the entire first print run—excepting review copies—and destroyed the books.
Bolton’s case would be significantly stronger if he had immediately self-published or posted the book on the Internet on April 28. He could argue that he relied on Knight’s assertion that the review was complete, and that he was entitled to do so. While prepublication review lawyers always like to insist on official written confirmation, I do not believe it is expressly required, and the government’s complaint supports that reading. The complaint states that “receipt of formal written notice of authorization is necessary to complete the prepublication process,” but then states, “Upon completion of that process, the staff of the Records Access and Information Security Management Directorate generally advises the submitter of a work in writing, either by email or letter, that the NSC’s classification concerns have been addressed and that the author is free to publish their work.” The word “generally” is noteworthy here, as it implies that there are instances when the prepublication review staff does not provide written notice, which undercuts the government’s assertion that “receipt of formal written authorization is necessary.” This reading is bolstered by the government’s claim that the NSC “does not act pursuant to any formal regulations governing its prepublication review process,” implying that prepublication review is more of an ad hoc affair at the NSC with no actual rules.
Accordingly, if Bolton had chosen to move forward and publish quickly without receiving the “customary pro-forma letter,” I think he would have had a decent chance of prevailing in this case. However, he did not, and as Shaffer did when the DIA intervened, Bolton was required to stop when the NSC informed him that it was re-reviewing the manuscript. He could still argue that he was being politically targeted, but he would have to argue it to a court. He could not just ignore it.
Sidebar: Can Bolton’s Lawyers See the Government’s Classified Evidence?
Before closing this analysis, it is necessary to address one recent development which is to my knowledge unique to this case. By pushing for prior restraint, the government may have given Bolton’s lawyers a benefit that other lawyers in prepublication review cases often lack, namely, the ability to see the government’s classified evidence and oppose it directly.
The case law on whether or not an author is entitled to have classified evidence shared with his lawyer in a prepublication review case is unsettled. In Shaffer’s case, the district court held that an agency’s refusal to process a lawyer for access to the classified information at issue could raise First Amendment concerns, but did not opine as to whether or not it did. In another case, the D.C. Circuit held that a district court could allow the author’s lawyer to participate in classified proceedings, but only if it could not “resolve the classification issue without the assistance of plaintiff’s counsel.” Neither one of these cases held that a judge must allow the author or his counsel to participate in the customary in camera proceedings, but neither one prohibited it either.
However, as Bolton’s counsel pointed out in a motion filed Friday night, none of the prepublication review cases where in camera, ex parte proceedings were allowed involved prior restraint. Bolton cites to one Supreme Court case which he claims “squarely held that there is ‘no place’ for ex parte procedures in the issuance of a prior restraint.” The actual holding of that case was a bit more nuanced than represented, however:
There is a place in our jurisprudence for ex parte issuance, without notice, of temporary restraining orders of short duration; but there is no place within the area of basic freedoms guaranteed by the First Amendment for such orders where no showing is made that it is impossible to serve or to notify the opposing parties and to give them an opportunity to participate.
However well-crafted this motion might have been last week, I predict it is doomed to failure now. It relied on the idea that Judge Lamberth might think that prior restraint is such a constitutionally onerous penalty that he should not consider it based on secret evidence. However, the judge did just that when he denied the government’s TRO motion on Saturday, even going so far as to opine that based on the classified evidence, “the Court is persuaded that Defendant Bolton likely jeopardized national security by disclosing classified information in violation of his nondisclosure agreement obligations.” With that argument settled, Bolton is in no better position than any other author to have his lawyers given access to the classified evidence. Moreover, since Judge Lamberth denied the government’s TRO motion, Bolton would have no ability to cite the judge’s reliance on in camera, ex parte evidence as grounds for an appeal.
With an adverse ruling being the likely outcome, the question for Bolton will become, was it worth it? He will have gone to all this trouble for no reward, and he still will have had to pay his attorneys. The only way he would feel like he came out on top was if he had done it all to expose and punish the president. However, if that were his motivation, then he could have simply told Congress what he knows. While the White House might try to insist that he would be required to submit Congressional testimony for prepublication review, the weight of authority would lie against such an imposition. By writing and publishing a book without completing the prepublication review process—however broken it might be—he will actually end up financially worse off than if he had simply written the manuscript and handed it to Congress.
Kel McClanahan is the executive director of National Security Counselors and an adjunct professor at the George Washington University Law School, where he teaches law of secrecy. He can be found on Twitter at @NatlSecCnslrs.
[image via Melissa Sue Gerrits/Getty Images]
This is an opinion piece. The views expressed in this article are those of just the author.