Individuals don’t lose their First Amendment rights by entering into government service. By now, that is black letter law. You wouldn’t know this, though, from the system of lifelong censorship that the government imposes on former intelligence agency employees and military personnel.

Under this system, known as “prepublication review,” millions of former public servants must submit their writing to the government for review prior to publication. And the government can censor their words with very few constraints. The review process frequently results in delays, sometimes severe ones, that prevent the people who know the most about how the government actually works from contributing to critical public debates. And the associated frustrations often cause authors to self-censor or sit out debates entirely.

On Tuesday, the Knight First Amendment Institute at Columbia University and the ACLU filed a lawsuit challenging the system, in its current form, on behalf of five former intelligence agency employees and military personnel.

Originally, prepublication review affected a small number of former government employees, mostly those who worked for the Central Intelligence Agency and had access to the nation’s most sensitive secrets. In the late 1970s, one CIA employee challenged the agency’s prepublication review regime on First Amendment grounds, but the Supreme Court dismissed his claim — without briefing or a hearing on the merits of the claim — in a single footnote.

Since then, the system has grown increasingly unwieldy, arbitrary, and far-reaching. Today at least 17 agencies administer some form of prepublication review. Review criteria are opaque and overbroad, and they differ across agencies. Authors have no meaningful recourse to contest unfavorable decisions. Censors have sweeping power to delay publication or to redact information regardless of whether it would reveal bona fide national security secrets. Moreover, the government appears to review manuscripts that are critical of intelligence agencies and government policies with greater scrutiny and more slowly.

This broken system violates the First Amendment because it infringes both the authors’ right to speak as well as the public’s right to hear them. This system also violates the Fifth Amendment because it fails to provide former employees with fair notice of what they can and cannot publish without prior review, and it invites arbitrary and discriminatory enforcement and redactions by government censors.

Even Congress recognizes that the system is broken. Nearly two years ago, it ordered the intelligence community to develop a new set of rules to govern prepublication review. The government has said that these rules are “forthcoming,” but Congress’s deadline has passed, and the director of national intelligence hasn’t given any indication of when the new rules will be published or implemented.

As former intelligence employees and military personnel, our clients bring unique and important perspectives to matters of great public concern. Yet they have been forced to choose between subjecting their writing to an unconstitutional censorship regime or risking sanction. Our clients take seriously their obligations to protect information essential to national security, and they understand the consequences, including criminal prosecution, of breaking the public’s trust. They believe, however, that the current system is broken and getting more broken every day.

One of our clients is Mark Fallon, who spent almost three decades with the Naval Criminal Investigative Service and two years at the Department of Homeland Security. When he wrote a book criticizing the Bush administration’s torture policies, the government sat on the manuscript for months.

The book, “Unjustifiable Means,” was based entirely on Fallon’s post-retirement research into the voluminous public record of the Bush administration’s use of torture. More than eight months after he submitted his manuscript, the government cleared it — but not until the ACLU and the Knight Institute wrote to six senators about the ongoing censorship, and various news outlets shined a light on the government’s extreme delay.

When it finally returned Fallon’s manuscript, the government required 113 separate redactions, several more than two pages long. Eager to finally add his voice to the ongoing public discussion, Fallon accepted the redactions, though they were seemingly arbitrary and diminished the book’s impact. In the end, in addition to delaying Fallon’s publication by many months, the government severely curtailed his ability to engage in the renewed public debate of torture sparked by then-presidential candidate Donald Trump.

Another of our clients — former CIA officer Mel Goodman — waited more than 11 months for his book, a memoir about his time in government that was critical of the agency, to clear prepublication review. When he finally got the manuscript back, the CIA insisted that he redact pages of information that was already in the public domain. He believes the redactions were largely intended to avoid embarrassing the agency, not to protect information that is legitimately classified.

Similarly, our client Richard Immerman — an academic who worked for less than two years in the Office of the Director of National Intelligence (ODNI) — had to wait nearly six months for a response after he submitted his book on the history of the CIA. He learned that the ODNI had sent his manuscript to the CIA, which conducted its own review and ended up demanding extensive redactions — all of which related to publicly sourced material, including material that the CIA itself had previously published. Although Immerman was able to successfully appeal the vast majority of these redactions, he had to spend another two months fighting with CIA censors to achieve that outcome.

Even former employees who have generally had smoother experiences with prepublication review, like our client Tim Edgar (also formerly of the ODNI), see the requirements of the process — what must be submitted, by whom, when, and for what purposes — as vague and confusing as well as ripe for abuse.

The system is so opaque that many others aren’t even aware of their obligations to submit. Our client Anuradha Bhagwati, for example, is a former Marine Corps officer who has written and spoken widely about her experiences in the military. It was only recently, however, that she learned that, as a former Defense Department employee, she is technically required to submit such works for review, even if she never had access to highly sensitive information while in government.

The government has an interest in protecting legitimate national-security secrets, but the current prepublication review system needs to be overhauled. Any system of prepublication review should apply only to those who had access to the most closely held government secrets and only to material reasonably likely to contain those secrets. The system should provide clear notice of what must be submitted and what standards will guide agency review. It should also clearly limit the discretion of government censors, include strict and definite time limits for completion of review, and require censors to explain their decisions. Finally, censors’ decisions should be subject to prompt review by the courts.

We are hopeful that our lawsuit will lead to a system in which former employees like our clients can participate in ongoing public debates without having to navigate a sprawling and burdensome censorship process that simply can’t be squared with the Constitution.