Prison Officials Cannot Censor Outgoing Mail Just Because They Dislike What an Inmate Says

Prison officials have broad authority to inspect and restrict mail, but that authority has limits, especially when the mail is leaving the prison.

In Grenning v. Key, No. 23-3018, 2026 U.S. App. LEXIS 26014 (9th Cir. Aug. 26, 2026), the Ninth Circuit reversed a ruling in favor of prison officials after concluding that the district court applied the wrong First Amendment standard to the censorship of an inmate’s outgoing writing. The appellate court held that because the case involved content-based screening of outgoing correspondence, the stricter rule from Procunier v. Martinez applied—not the much more deferential standard from Turner v. Safley that governs most prison regulations.

That distinction was critical. Under Turner, prison officials generally receive substantial deference so long as a restriction is reasonably related to a legitimate penological interest. Under Martinez, however, officials must make a much stronger showing before they can censor outgoing correspondence. The Ninth Circuit found that the district court gave prison officials too much deference by analyzing the case under Turner and, in doing so, failed to apply the stronger First Amendment protections that continue to govern mail leaving a prison.

Prison Officials Blocked an Inmate’s Writing From Leaving the Prison

Neil Grenning was incarcerated at Airway Heights Corrections Center in Washington and participated in a creative writing program sponsored by Eastern Washington University.

The program encouraged incarcerated participants to develop their writing and submit their work for possible publication in InRoads, a university literary journal.

Grenning wrote a short story titled Dark Room. The story dealt largely with a father’s relationship with his daughter and memories of his deceased wife. One passage briefly described a sexual encounter between the husband and wife.

Grenning submitted the story through the prison writing program.

Ann Wise, the prison’s Community Partnership Program Coordinator, stopped the submission and returned it to Grenning. She said the story was inappropriate for publication because of its “sexual content.”

The university publication expressly warned contributors that its audience included children and other sensitive readers, but it also stated that mature themes and adult language could be accepted at the editor’s discretion. Previous issues had included references to rape, teen pregnancy, drug use, suicide, arson, and death. The prison writing program itself assigned books containing sexual assault and other mature material.

More importantly, prison officials later acknowledged that Grenning’s story was not graphic enough to violate the prison’s own restrictions on sexually explicit material. Yet Wise stopped the story before the university’s editors could even decide whether they wanted to publish it.

The Prison Superintendent Said the Real Concern Was “Negative Publicity”

Grenning eventually mailed the story directly to Eastern Washington University and complained to prison superintendent James Key. Key backed Wise’s decision. But Key’s explanation became one of the most damaging facts in the case.

He told Grenning that the program coordinator was responsible for making sure that nothing left the institution through the volunteer program that could cause “negative publicity” for the prison or Department of Corrections.

Key also warned Grenning that if he again sent correspondence directly to a program volunteer without approval, he would be removed from the writing program.

Grenning sued, arguing that prison officials violated the First Amendment by censoring his outgoing writing and retaliating against him when he tried to send it himself.

The district court sided with the prison. The Ninth Circuit reversed.

The District Court Used Turner v. Safley, And the Ninth Circuit Said That Was the Wrong Test

This is the most important part of the decision. The district court analyzed Grenning’s claim under Turner v. Safley, the standard normally used when prisoners challenge prison regulations that affect constitutional rights.

Turner is highly deferential to prison administrators. Under Turner, a prison regulation generally survives constitutional review if it is reasonably related to a legitimate penological interest. Courts look at factors such as whether there is a rational connection between the rule and the asserted correctional interest, whether the prisoner has alternative ways to exercise the right, and what effect accommodating the right would have on prison operations.

That framework usually favors the prison. The district court applied Turner and accepted two asserted interests for blocking Grenning’s story: preventing inmates from using a prison educational program to distribute sexual material to children and avoiding negative publicity for the institution.

The Ninth Circuit said the district court started from the wrong place. The case was not simply about a prison program rule.

It was about a prison official screening outgoing written expression based on what the writing said and refusing to allow it to leave the institution.

That put the case squarely within Procunier v. Martinez, which specifically governs censorship of outgoing prisoner correspondence.

Why Martinez Gives Outgoing Mail More Protection

The Ninth Circuit explained that the Supreme Court has intentionally treated outgoing correspondence differently from incoming mail and many other prison regulations.

Martinez held that censorship of outgoing prisoner mail is constitutional only if two requirements are satisfied.

First, the restriction must further an important or substantial governmental interest that is unrelated to suppressing expression. The recognized interests include prison security, order, and rehabilitation.

Second, the restriction cannot sweep more broadly than necessary to protect that interest.

That is a substantially tougher standard than Turner. And the reason for the difference is practical: outgoing mail generally poses less danger to prison security than material coming into a prison.

The Supreme Court later adopted Turner’s more deferential test for many prison regulations, but the Ninth Circuit emphasized that the Supreme Court never eliminated Martinez when it comes to outgoing mail.

In Thornburgh v. Abbott, the Supreme Court applied Turner to incoming publications but expressly preserved Martinez for outgoing correspondence, explaining that the security implications of outgoing mail are “categorically lesser” than those created by material entering a prison.

The Ninth Circuit pointed out that it has repeatedly followed that distinction. Its prior cases continue to apply Martinez to outgoing inmate mail, while Turner generally governs incoming publications and ordinary prison regulations.

In other words, Turner did not replace Martinez across the board. Martinez still controls when prison officials censor outgoing correspondence because of its content.

The Court Found Grenning’s Case Closely Resembled Martinez

The Ninth Circuit did not view this as some new extension of First Amendment law. It said the facts closely resembled Martinez itself.

In Martinez, prison officials screened outgoing personal mail and could refuse to send letters when they disapproved of the content. One of the rules struck down by the Supreme Court allowed officials to censor correspondence considered lewd or obscene.

That was essentially what happened to Grenning. Wise reviewed his outgoing writing, disapproved of its sexual content, and prevented it from being transmitted.

The Ninth Circuit emphasized that her decision was based on a vague and discretionary judgment that the story was “inappropriate for publication,” not on a demonstrated security threat, threat to institutional order, or concrete rehabilitative concern.

The court saw that kind of discretionary, content-based judgment as exactly the type of censorship Martinez was designed to prevent.

“Negative Publicity” Was a Major Problem for the Prison

The prison’s own explanation made the constitutional problem worse. Superintendent Key said Wise was supposed to ensure that nothing left the prison that could generate negative publicity.

The Ninth Circuit compared that rationale directly to the censorship condemned in Martinez. Prison officials cannot censor outgoing speech merely because the speech is embarrassing, critical, unwelcome, or likely to make the institution look bad.

Martinez specifically rejected censorship designed simply to eliminate unflattering or unwelcome opinions. That made the “negative publicity” justification particularly difficult for the prison to defend.

The court explained that Grenning had a clearly established right not to have nonthreatening outgoing expression blocked merely because officials disapproved of the content or feared the public’s reaction to it.

That language is important because it extends well beyond a creative writing program.

It goes directly to the ability of prisoners to communicate with people outside the institution.

The Court Also Rejected a Vague Claim of “Rehabilitation”

The prison also relied on rehabilitation. Rehabilitation is unquestionably a legitimate correctional interest. But the Ninth Circuit said prison officials cannot simply invoke the word “rehabilitation” and end the constitutional inquiry.

There must be a sufficiently close connection between the particular restriction and the actual rehabilitative goal being asserted.

The court warned that if broad claims such as “accountability,” “appropriate communication standards,” or “appropriate boundaries” were enough by themselves, prison officials could censor almost any outgoing expression by claiming that they considered it inconsistent with rehabilitation.

That would effectively swallow the First Amendment rule. The court therefore required something more concrete. And on this record, the prison had very little.

Officials admitted that Grenning’s story did not violate the institution’s actual sexually explicit material restrictions. There was no evidence that allowing the story to be submitted would threaten prison security, create disorder, encourage sexual misconduct, or otherwise undermine the prison’s rehabilitative mission.

The Prison’s Own Mail Policy Hurt Its Position

The Ninth Circuit also looked at the prison’s own policies. Those policies allowed mail to be inspected and read for legitimate reasons such as preventing contraband, threats to institutional order or security, and criminal activity.

But they also contemplated that outgoing inmate mail would bear a notice essentially advising the recipient that the material had been mailed by a prisoner and that its contents were uncensored.

The policy governing sexually explicit material also required significant or repeated instances of prohibited sexual content before material could be rejected.

Grenning’s story did not meet that standard. And there was no indication that the story was referred to the prison’s Publication Review Committee, even though the policy provided for further review and allowed otherwise restricted material to be approved for artistic, educational, or medical purposes.

The Ninth Circuit viewed those facts as additional evidence that the decision may have been based less on an established correctional rule and more on the personal judgment of the officials involved.

Calling Censorship “Screening” Does Not Change What It Is

The prison also argued that the approval process was merely a form of screening connected to the writing program.

The Ninth Circuit rejected that characterization. The court made clear that constitutional rights cannot be avoided by changing labels.

If a prison official refuses to transmit outgoing written expression because of its content, calling the process “screening,” “prescreening,” or an “approval requirement” does not remove the First Amendment issue.

The court summed up the point sharply: the government cannot defeat a First Amendment claim by using different terminology. What matters is the substance of what happened.

Grenning wrote something intended for an outside recipient. A prison official read it, objected to the content, and refused to allow it to leave through the established program. That is why Martinez applied.

The Ninth Circuit Said the Prison Would Have Problems Even Under Turner

The court went even further. It said that even if the district court had been correct and Turner applied, the prison still would not have been entitled to summary judgment.

The first Turner factor requires a valid, rational connection between the restriction and a legitimate governmental interest. The Ninth Circuit found serious problems there. Avoiding negative publicity was plainly not a legitimate basis for suppressing speech.

And merely pointing to “sexual content” was not enough either, particularly when the prison admitted the story did not violate its own sexual-material policy.

The court also found no evidence connecting publication of the story to sexual misconduct, threats to institutional security, or other concrete correctional harms.

So even under the more prison-friendly Turner test, the Ninth Circuit said the government had not demonstrated the necessary rational connection between censoring Grenning’s story and the interests it claimed to be protecting.

What This Means for People in Prison

Grenning is important because it reinforces a distinction that many prisoners—and even many prison officials—may not fully appreciate:

Incoming prison mail and outgoing prison mail are not governed by the same constitutional standard.

Prisons receive substantial deference when regulating material entering an institution because incoming publications and correspondence can create direct security concerns.

Outgoing communication is different. When prison officials stop nonthreatening written material from leaving a prison because they dislike the content, fear embarrassment, or simply believe the speech is inappropriate, Martinez requires a much stronger justification.

The prison must tie the restriction to a real and substantial interest involving security, order, or rehabilitation. It cannot rely on vague labels. It cannot rely on public-relations concerns. And it cannot convert content-based censorship into something constitutional merely by calling it an approval process.

That could matter when incarcerated people are communicating with family members, journalists, advocacy organizations, publishers, government agencies, consultants, or others outside the prison.

The case does not mean prisons lose control over outgoing correspondence. Threats, criminal communications, contraband activity, escape plans, harassment, and legitimate security concerns can still justify restrictions.

But when the real justification is essentially “we do not want this leaving the prison because we do not like what it says,” Grenning provides strong authority that the First Amendment requires much more.