Published by The Lawfare Institute
in Cooperation With
The U.S. government filed a civil suit on June 17 against former National Security Adviser John Bolton. It primarily seeks an injunction against the planned June 23 publication of Bolton’s book, “The Room Where It Happened: A White House Memoir,” and a “constructive trust” that would give the United States the right to all of Bolton’s profits from the book. The case has been assigned to Judge Royce Lamberth of the U.S. District Court for the District of Columbia.
This post explains the case and offers our initial thoughts. The big news to us about the government’s case is that it’s weaker than we expected. We should emphasize, however, that these views are preliminary and incomplete. The case implicates a complex and in some ways unsettled area of law.
The Relevant Contract (Nondisclosure Agreement) Terms
The most important thing to understand about the case is that the government is suing Bolton for a breach of contract—two contracts, in fact. As the government’s complaint describes, the contracts in question are nondisclosure agreements (NDAs) that Bolton signed on April 5, 2018, when he entered government service as national security adviser. These NDAs are included as attachments to the complaint. (Correction: Actually, there were three such contracts, because Bolton signed two separate, substantively identical SF-4414s, corresponding to two different sets of Specialized Compartmented Information programs to which he was afforded access. Our references in this post to Bolton’s SF-4414 apply in kind to both of those agreements.)
The first NDA, Standard Form 312, contains obligations Bolton assumed as a condition of obtaining access to classified information generally, that is, a “security clearance.” Two are pertinent here. First, Bolton agreed that he must never “divulge classified information to anyone” unless either he “officially verifie[s] that the recipient has been properly authorized by the United States Government to receive it” or he has received “prior written notice of authorization [to divulge it] from the United States Government” entity responsible for its classification. Second, he agreed that if he’s “uncertain about the classification status” of any information, he must “confirm from an authorized official that the information is unclassified before [he] may disclose it” to anyone not authorized to receive it. We’ll refer to this second obligation as the SF-312 prepublication review requirement, although, as we explain below, it’s actually something less than that—it’s more like a simple “confirmation” requirement.
The second NDA, Standard Form 4414, set the conditions of Bolton’s access to a more sensitive subset of “Special Access Programs” classified information, commonly known as sensitive compartmented information (SCI). It requires Bolton to submit to a security review by the National Security Council (NSC) “any writing ... that contains or purports to contain any SCI or description of activities that produce or relate to SCI or that I have reason to believe are derived from SCI, that I contemplate disclosing to any person not authorized to have access to SCI or that I have prepared for public disclosure.” It further provides that he “will not disclose the contents of such preparation with, or show it to, anyone who is not authorized to have access to SCI until I have received written authorization from the Department or Agency that last authorized my access to SCI that such disclosure is permitted.” We’ll call this the SF-4414 prepublication review requirement.
The NSC’s Prepublication Review of Bolton’s Book
The government’s complaint tells a detailed story of the prepublication review process for Bolton’s book. The complaint and the attachments submitted along with it—including correspondence between Bolton’s legal team and the government—help piece together the narrative of what happened up to the present day.
After leaving the government in September 2019, Bolton drafted a book in which he was, according to his lawyer, Chuck Cooper, “careful ... to avoid including any classified information.” Nonetheless, on Dec. 30, 2019, Cooper submitted a manuscript of the book to Ellen Knight, a career NSA official who 12 days earlier had become the NSC senior director for records access and information security management—the entity within the NSC that ordinarily conducts classification and prepublication reviews. Knight’s office began reviewing the manuscript, which was more than 500 pages long.
Twenty-four days later, on Jan. 23 of this year, Knight informed Cooper that “[b]ased on a preliminary review, the manuscript appears to contain significant amounts of classified information,” and that the “manuscript may not be published or otherwise disclosed without the deletion of this classified information.” Thus began a three-month-long, “iterative” process between Bolton and Knight in which they negotiated many changes to the manuscript in order to ensure that, in Knight’s view, it wouldn’t contain any classified information. Bolton and Knight met for numerous hours, and had many written exchanges, as described in paragraphs 41-45 of the complaint.
Finally, as the government acknowledges, on April 27, Knight “had completed her review and was of the judgment that the manuscript draft did not contain classified information.” According to a letter from Cooper to National Security Council Legal Adviser John Eisenberg, Knight told Bolton on that day that she had made her final edit. Bolton asked her when he would receive a letter confirming that the prepublication review was over. Knight replied, according to Cooper, that it might be transmitted as early as that day, but she also noted there were “some internal process considerations to work through” at NSC. (Update: Less than an hour after we published this post, the government disclosed that on Tuesday, presumably after the Justice Department filed its complaint, the NSC sent Bolton a marked copy of the current version of the manuscript identifying passages that Michael Ellis, the new NSC senior director for intelligence programs, has determined, “based on [his] initial review, appeared to contain classified information.”)
According to the government, at that point the current national security adviser, Robert O’Brien, became “concerned that the manuscript still appeared to contain classified information.” O’Brien allegedly requested Michael Ellis to commence an additional review of the manuscript. Ellis currently serves as the NSC’s senior director for intelligence, a position he assumed on March 1. (Ellis previously worked as an aide to the House Intelligence Committee’s ranking Republican, Devin Nunes, and then as a deputy to Eisenberg when Bolton was national security adviser; Ellis was also allegedly involved in the decision to place the transcript of President Trump’s controversial July 25, 2019, call with Ukrainian President Volodymyr Zelensky in the White House’s most secure classified computer system.)
So why did Ellis review the manuscript after Knight thought it no longer contained classified information? According to the complaint, Ellis “is in a position to know intelligence information and internal foreign policy deliberations and developments that others of the NSC staff”—including Knight—“do not know.” On May 2, five days after Bolton and Knight had last spoken, Ellis began a new investigation of whether the book contained classified information. On May 6, Bolton emailed Knight: “Hope springs eternal—any news on the letter?” The next day, Knight replied tersely that she didn’t have any new information, that “[t]he process remains ongoing,” and that she would “reach out as soon as there is an update to provide.” Judging from the material provided by the government in the complaint, that appears to have been the last message Bolton received from the NSC for more than a month.
Finally, on June 8, Eisenberg wrote to Cooper to inform him that the current draft of the book manuscript “still contains classified information.” Eisenberg added that the NSC would provide proposed redactions to Cooper by June 19 (that is, this Friday) and that, in the meantime, Bolton could not publish or disseminate any portion of the manuscript. The next day, Ellis allegedly completed his classification review, but the NSC has not yet identified to Bolton or to Cooper the classified information that allegedly remains in the book.
Last Wednesday, June 10, Cooper responded to Eisenberg that Bolton had fully discharged his duties under the NDA, that the manuscript doesn’t contain any classified information, and that the White House is thus seeking to block publication for “purely political reasons” and using alleged national security concerns as a “pretext to censor” Bolton’s book. Cooper added that the book has been printed, bound and shipped to distributors, and that Bolton lacks any authority to stop Simon & Schuster from making the book available to the public next Tuesday, June 23—six days from now.
The Breach-of-Contract Suit
On June 17, the Justice Department sued Bolton, in the U.S. District Court for the District of Columbia, on behalf of the United States for breaches of the two NDAs. The government asks the court to impose four primary forms of relief for the alleged breaches:
(i) enjoining Bolton himself from “taking any steps towards publicly disclosing the information in The Room Where it Happened without first obtaining written permission from the United States through the prepublication review process”;
(ii) requiring Bolton, “insofar as he has the authority to do so, to further delay the release date of The Room Where it Happened until completion of the prepublication review process; and to instruct or request his publisher, insofar as he has the authority to do so, to take any and all available steps to retrieve and dispose of any copies of The Room Where it Happened that may be in the possession of any third party” (emphasis added);
(iii) imposing a “constructive trust” that would give the United States the right to “all monies, gains, profits, royalties, and other advantages that Defendant and his agents, assignees, or others acting on his behalf have derived, or will derive, from the publication, sale, serialization, or republication in any form, including any movie rights or other reproduction rights, of The Room Where it Happened” (see paragraph 12 of both NDAs);
(iv) declaring that, pursuant to Federal Rule of Civil Procedure 65(d)(2), the court’s order prohibiting disclosure of the information in “The Room Where It Happened” will “bind Defendant’s agents and other persons who are in active concert or participation with Defendant or his agents, … including Simon & Schuster, Inc. and other such persons in the commercial distribution chain of Defendant’s book,” when they receive formal notice of the order, even though the Justice Department hasn’t named them as parties in the suit.
Notably, the government has not (yet) asked for a temporary restraining order or a preliminary injunction against publication. It may already be too late. The book has been distributed and will be widely sold next Tuesday. The New York Times, among other outlets, has a copy and is reporting on its contents and reviewing it. Bolton himself previewed the book in the Wall Street Journal on June 17. This all means that the focus of the case will almost certainly be on whether the government can impose a constructive trust and obtain Bolton’s profits. (Update: Less than an hour after we published this post, the government did file a motion for Judge Lamberth to issue a temporary restraining order—one that would prohibit not only Bolton himself, but also Simon & Schuster and “[c]ommercial resellers further down the distribution chain, such as booksellers,” from selling or distributing “The Room Where It Happened.” The government has asked for a hearing on its motion on Friday, June 19, four days before the publication date and has offered to provide Lamberth with evidence supporting its motion that he would review in camera and ex parte.)
Some Views on the Merits
The government’s requests for relief rest on its claim that Bolton has committed two types of breaches of contract. One breach involves Bolton’s alleged disclosure of classified information in the book and in prior circulation of the manuscript. The second concerns Bolton’s violations of his alleged preclearance obligations, independent of whether the book contains classified information. We’ll discuss these theories separately.
1. Alleged Breach Based on Disclosure of Classified Information
In order to prevail with respect to its first theory of alleged breach—and certainly in order to have any shot at having Judge Lamberth enjoin the book’s publication on that ground—the government must carry its burden of demonstrating to the court that the book contains properly classified information, even though Ellen Knight concluded, after a long and arduous review, that no classified information remained in the edited manuscript.
Under the executive order that establishes the basis for classifying government information, the government would have to demonstrate, at a minimum, that the unauthorized disclosure of information in Bolton’s book “reasonably could be expected to result in damage to the national security,” where “national security” is defined to include “the national defense or foreign relations of the United States.” This is not exactly a precise standard. In the leading case on the question of how a court should review a classification decision in this context, McGehee v. Casey, the U.S. Court of Appeals for the D.C. Circuit stated that “[s]tanding alone, such a classification standard might be excessively vague.” The court in McGehee was satisfied that the standard was not excessively vague, however, because the CIA (the agency involved in that case) had prescribed more specific criteria to govern its prepublication review practice.
The court in McGehee ruled that judicial review of agency prepublication decisions about proper classification should be de novo, albeit with some deference to any “reasoned and detailed … explanations” that the agency offers for its decisions:
While we believe courts in securing such determinations should defer to [agency] judgment as to the harmful results of publication, they must nevertheless satisfy themselves from the record, in camera or otherwise, that the [agency] in fact had good reason to classify, and therefore censor, the materials at issue. Accordingly, the courts should require that [agency] explanations justify censorship with reasonable specificity, demonstrating a logical connection between the deleted information and the reasons for classification. These should not rely on a “presumption of regularity” if such rational explanations are missing. We anticipate that in camera review of affidavits, followed if necessary by further judicial inquiry, will be the norm. …
[C]ourts must assure themselves that the reasons for classification are rational and plausible ones. (Emphasis added.)
Importantly, McGehee involved review of a classification decision in a suit brought by a former CIA employee who alleged that the CIA erroneously deleted items from the material being reviewed during the preclearance review process. In a footnote, the court distinguished that sort of case from one in which the agency itself seeks “an injunction against publication of the censored items,” and opined that in the latter sort of case the government “would bear a much heavier burden.” (In a concurrence, Judge George MacKinnon challenged the court’s dicta about a heavier burden when the government is the moving party. No court since then has discussed the issue.)
In the Bolton case, whether Lamberth applies the sort of review prescribed in McGehee itself or imposes an even heavier burden on the government (as the McGehee footnote suggested), the government might have a hard time meeting its burden before the June 23 publication date—especially in light of Knight’s extensive vetting of Bolton’s manuscript, the many changes Bolton made to it in response, and Knight’s “judgment [on April 27] that the manuscript draft did not contain classified information.” Whether the government can persuade Lamberth “that the reasons for classification are rational and plausible ones” may depend on what sort of showing Ellis makes this Friday, when he’s scheduled to explain to Bolton how the manuscript continues to contain classified information. (Update: In its new motion for a temporary restraining order, the government has offered four public and two classified declarations, including one of each from Ellis, in support of its claim that the book contains classified information. We can’t assess the possible strength of that assertion at this point, because the supporting details appear only in the classified declarations.)
A final note on this theory of breach. The government alleges not only that the printing, distribution and imminent publication of the book would violate Bolton’s duty not to disclose classified information. It also alleges that Bolton already violated his obligations under SF-312 by sharing classified information “numerous times” prior to printing and publication, during the prepublication review, with his attorney (Cooper), his publisher (Simon & Schuster), and “numerous acquaintances and friends, and members of the news media.”
The government may be helped in its argument that Bolton committed unlawful prepublication disclosure by the fact that Bolton allegedly circulated this material to persons unauthorized to view classified information before the end of the long iterative process with Knight that excised a good deal of allegedly classified information. That said, Knight and others were well aware that Bolton’s attorney had seen the early manuscript, and nothing in the record suggests that anyone complained at the time about this standard practice. It’s also unclear whether the government is seeking any remedy for the alleged prepublication sharing of classified information apart from a simple declaration of wrongdoing.
2. Alleged Breach by Publishing the Book Before the Prepublication Review Is Complete
The government also faces hurdles under its second theory of breach—that Bolton has violated his prepublication review duties.
Start with Standard Form 312. SF-312 has a strict prohibition on the disclosure of classified information. But it imposes a quite soft obligation to preclear publications with the government. The key sentence reads: “I understand that if I am uncertain about the classification status of information I am required to confirm from an authorized official that the information is unclassified before I may disclose it ...” (emphasis added).
Bolton will likely argue that he never harbored such uncertainty—not when he first submitted the manuscript and definitely not after Knight’s clearance—her “confirm[ation]” that the redacted draft had no classified information—at the end of the extensive iterative review. He’ll likely also argue that the SF-312 preclearance obligation, unlike the SF-4414 obligation (see below), doesn’t require written approval from the government. It only requires “confirmation from an authorized official that the information is unclassified”—something that Knight reportedly gave Bolton on April 27. Eisenberg wrote in his June 11 letter to Cooper that Bolton “is well aware that the manuscript still contains classified information,” in part because Bolton himself allegedly classified some of the information when he was in government. If true, this would be the government’s strongest argument under SF-312. But Bolton might respond that whatever he did in government, he satisfied his contractual obligations to allay any possible uncertainty about classification after Knight confirmed to him that the current version of the manuscript contains no classified information.
Now let’s turn to SF-4414. It contains a much more robust preclearance obligation than SF-312. It requires Bolton to submit to preclearance review “any writing” that, inter alia, describes “activities that … relate to SCI” (emphasis added), and obligates him not to disclose or publish the writing until he receives written authorization from the proper government agency, which Bolton has not yet received. It might therefore appear, at first glance, as though the government would be able to obtain both an injunction and a constructive trust, relying on the Supreme Court’s 1980 decision in Snepp v. United States. On closer inspection, however, the government has some significant challenges to overcome in order to obtain those remedies.
First, the “relates to SCI” criterion in the NDA, which has (to the best of our knowledge) never been litigated, is notoriously vague and quite possibly too vague to justify a prior restraint.
Second, there’s reason to doubt that the information in “The Room Where It Happened”—which Knight concluded contained no classified information—“relates to” SCI. Indeed, neither Eisenberg’s letters to Cooper nor the government’s complaint alleges overtly that the manuscript contains any information relating to (or derived from) SCI, let alone any SCI itself. Eisenberg’s June 11 letter, for instance, claimed that the book “still contains classified information” but did not allege that it contains SCI. As for the complaint, it says that Ellis has access to “extremely sensitive intelligence reports” and implies that he might have spotted relevant information that Knight missed. Perhaps that included some SCI. But probably not. The complaint throughout treats “classified information and sensitive compartmented information” separately (see, for example, paragraphs 13 and 71) but doesn’t ever allege that the manuscript contains SCI. And the carefully worded prayer for relief only seeks remedies related to Bolton’s alleged publication of classified information. Although the claims for relief include references to SCI, the alleged legal violations and harms to the United States are limited to disclosures of classified information.
Thus it appears that the government neither believes nor alleges that the book contains sensitive compartmented information. And that’s especially important because Section 5 of SF-4414 itself specifies that the prepublication review the NDA requires is not for purposes of classification review broadly, nor for redacting information “related to” SCI, but only to ensure that SCI itself is not disclosed. It reads: “I understand that the purpose of the review described in paragraph 4 is to give the United States a reasonable opportunity to determine whether the preparation submitted pursuant to paragraph 4 sets forth any SCI” (emphasis added). If, as appears to be the case, the government does not believe Bolton’s manuscript contains any SCI, then Lamberth might well conclude that the NSC lacks a basis to deny Bolton the necessary “written authorization” to publish. And if the judge concludes that NSC is required to provide such authorization, then the publication of the book likely would not violate Bolton’s obligations under SF-4414. (Update: In a declaration attached to the motion for a temporary restraining order that the government filed just after we published this post, Michael Ellis attests—see ¶ 19—that the soon-to-be published version of “The Room Where It Happened” does, indeed, contain at least some Specialized Compartmented Information–even though Knight and her NSC review team presumably didn’t find any such SCI in the final manuscript at the end of their painstaking, months-long review. Whether Ellis can persuade Lamberth that the book does, indeed, contain SCI might be one of the most important considerations in the government’s current effort to obtain a temporary restraining order enjoining publication–and any further efforts to obtain a constructive trust over Bolton’s profits from the book.)
For these reasons, it’s very far from clear that Lamberth will conclude that Bolton has violated his prepublication obligations under the NDAs. The primary disputed issue will probably be whether Bolton has any uncertainty about whether any material in the book is classified, particularly in light of Knight’s confirmation that the manuscript contains no classified information.
And that’s before we even get to the First Amendment issues related to prior restraints.
There are many complexities and uncertainties about application of the First Amendment here, and many relevant lower court decisions, discussion of which we can save for another post, if necessary. For now, we’ll simply offer a few thoughts on why the Supreme Court’s decision in Snepp might be less helpful to the government than many observers believe, even if the government were able to show that Bolton will breach a requirement of waiting to publish until completion of the prepublication review. There’s a powerful presumption against prior restraints under the First Amendment. Some people read Snepp to be a categorical exception to this presumption in the context of violations of prepublication review requirements. But that’s almost certainly an overreading of Snepp.
The prepublication review obligation at issue in Snepp was different from Bolton’s obligations under SF-312 and SF-4414. Snepp, a former CIA employer, agreed not to “publish ... any information or material relating to the Agency, its activities or intelligence activities generally, either during or after the term of [his] employment … without specific prior approval by the Agency.” Yet Snepp nonetheless published a book about the CIA without even submitting it for prepublication review.
The government sued, seeking a declaration that Snepp breached his contract, an injunction to require Snepp to submit future writings for prepublication review (in effect, a prior restraint), and an order imposing a constructive trust on Snepp’s profits. Snepp argued that the contract requiring preclearance review was an illegal prior restraint. The Supreme Court rejected this argument in passing in a footnote with practically no analysis. All it said was: “The Government has a compelling interest in protecting both the secrecy of information important to our national security and the appearance of confidentiality so essential to the effective operation of our foreign intelligence service. The agreement that Snepp signed is a reasonable means for protecting this vital interest.”
Commentators ever since have rebuked Snepp’s cursory treatment of the First Amendment issue—especially because the Supreme Court decided the issue summarily, without full briefing or oral argument. These considerations will be relevant if Bolton’s case reaches the Supreme Court. But even before then, Snepp doesn’t guarantee that the government will get its requested injunction against Bolton if Judge Lamberth concludes that he violated his preclearance obligations.
The question in Snepp was framed as whether the preclearance contract itself was a prior restraint. Although the Supreme Court did approve a judicial injunction against future disclosures without prepublication review, it did so in the context of an employee who simply refused to invoke the contractual process at all. Bolton, by contrast, submitted his manuscript for preclearance review, went to elaborate efforts to comply with the process for months, made numerous redactions of information when requested by Knight, and was eventually told by her that the heavily redacted version of his book contained no classified information. Assuming, contrary to the points made above, Bolton would breach his preclearance obligations nonetheless by publishing the book without a final NSC sign-off, the government is asking for an injunction against Bolton (and through Federal Rule of Civil Procedure 65, effectively against his publishers, distributors and booksellers, too) to prevent publication of a book that has been printed and is slated to go on sale next week.
This posture would present the First Amendment question in a far different light than Snepp. The First Amendment’s influence should weigh more heavily in Bolton’s case than it did in Snepp—and, at the very least, it will likely inform all the uncertainties we outlined above concerning whether Bolton has breached his contracts.
This piece has been cross-posted to Just Security.
Editor's note: A previous version of this piece misstated the number of nondisclosure agreements signed by Bolton. It has been updated to incorporate developments that took place shortly after publication.