Month: December 2025

The Department of Defense in late November 2025 announced that it would investigate U.S. Sen. Mark Kelly, a retired Navy captain and NASA astronaut, for what Secretary of Defense Pete Hegseth has called seditious behavior. The threat of investigation came after Kelly and five other Democrats, all with military backgrounds, released a video reminding U.S. service members they can disobey illegal orders issued by the Trump administration.

“No one has to carry out orders that violate the law, or our Constitution,” the lawmakers said, without specifying the orders the U.S. service members may have received. “Know that we have your back … don’t give up the ship.”

In response to the video, President Donald Trump accused the lawmakers of “seditious behavior” that could be “punishable by death.”

Sedition is a federal crime, but as a military law scholar who served as a judge in the U.S. Air Force, I believe the Democratic lawmakers articulated a correct view of military law. That is, service members subject to the Uniform Code of Military Justice have a duty to not obey unlawful orders.

There are several unique features to military law that have no analog to civilian criminal law, and if Kelly were court-martialed he would be deprived of several fundamental constitutional rights.

Military justice

In a civilian criminal trial the government normally has the burden of proof on all matters. But in a court-martial, a service member who argues that an order is unlawful has the burden of proving its unlawfulness. And the Supreme Court, in its 1827 opinion in Martin v. Mott, gave this view some credence, arguing that the president, as commander in chief, should not be questioned during a national emergency.

Second, ordinary citizens are protected by a constitutional requirement that the prosecution must convince all jurors of the defendant’s guilt beyond a reasonable doubt. A court-martial has only a two-thirds threshold to establish guilt. And the jurors – called members – are not the accused service member’s peers.

Indeed, the court-martial members are military personnel who outrank the accused service member and are picked to serve by senior commanding officers. Military judges are also uniformed officers and, like the rest of the military, are subject to the chain of command.

At times, senior officers have inserted themselves into the military justice system and tried to direct a court-martial to convict an accused service member. This has created the problem of unlawful command influence, the improper use of superior authority to interfere with the court-martial process.

Kelly is still theoretically subject to the Uniform Code of Military Justice and could be court-martialed because he is a military retiree. This concept of a lifetime military jurisdiction did not exist when the Constitution was instituted in 1789. It came into existence during an emergency session of Congress in 1861.

The Supreme Court has never held that lifetime jurisdiction is constitutional. But in 2022 the U.S. Court of Appeals for the District of Columbia did, in a 2-1 decision.

It reasoned that if the Constitution’s creators had thought such a jurisdiction were a threat to the republic, they would have prohibited it. The dissenting judge in that case pointed out the frightening possibility of a president using the Uniform Code of Military Justice to curb free speech.

Lines of defense

Kelly is different than an ordinary retiree, and this case is bigger than a single senator. That’s because it goes to the heart of what the Constitution’s framers intended by preserving liberty through a republican form of government.

In 1648, Oliver Cromwell, who had become a military dictator over England, used the army to curb the Magna Carta – a revolutionary basic rights document dating to 1215 – and the ability of Parliament to debate matters and pass laws. The Constitution is designed to prevent anything coming close to such an occurrence.

So, what would Kelly’s defense likely be, other than that he exercised free speech and gave a correct recitation of the law?

Kelly’s first defense might be that under the Constitution, the president, as commander in chief, has no power to court-martial or otherwise administratively penalize him. Doing so would diminish Congress’ authority.

In 1974, the Supreme Court determined in Schlesinger v. Reservists Committee that although the Constitution prohibits a member of Congress from holding a position in the executive branch, citizens had no standing to sue in the federal courts to prevent this from occurring. Taken literally, the clause means that no member of Congress could hold a military commission and be beholden to the commander in chief, since this would erode Congress’ independence and authority.

Kelly’s second defense could be that after the Constitution and statutory law, the military law is governed by tradition, or the military’s own past practices, which used to be referred to as “lex non scripta.”

American history is replete with retired officers criticizing presidents or even joining in hate groups that accused a president of being beholden to subversive interests. Past presidents have ignored these men.

They include George Van Horn Moseley, who sided with pro-Nazi groups and accused President Franklin Roosevelt of being a communist. Retired generals Albert Coady Wedemeyer and Bonner Fellers formed organizations that undermined Presidents Harry Truman and Dwight Eisenhower.

None of these men were court-martialed or administratively penalized.

Finally, Kelly could argue in federal court that the military has no jurisdiction over him because of the issue of unlawful command influence. One only needs to look at Hegseth’s statements in the case to see the specter of this problem in regard to Kelly.

When Congress formulated the Uniform Code of Military Justice, it criminalized unlawful command influence. But as military law scholar Rachel VanLandingham has pointed out, no person has ever been prosecuted for violating the prohibition.

Kelly could argue that there are no safeguards in his case to ensure a fair hearing and that the case should move from military courts to federal courts. The federal judge assigned the case can then ponder whether siding with the administration’s claims is a step toward establishing a Cromwellian future and away from the Constitution’s protection of a republican form of government.

Of course, Congress could put a stop to any persecution of Kelly by informing the president that he is acting contrary to the Constitution and explaining to do so is a high crime or misdemeanor.

During the Vietnam War, scholar Robert Sherrill said that “military justice is to justice what military music is to music.” In the past, military justice has been able to accomplish fair trials of military members, but it is dangerously open to influence by military leaders, all the way up to the commander in chief.

If there is to be an exercise in accountability for Kelly, it could more fairly be administered through a real constitutional analysis conducted by the independent federal judicial branch – or through a congressional intervention. Without either occurring, we may as a nation find ourselves a closer step toward a Cromwellian future.

Joshua Kastenberg does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

​Politics + Society – The Conversation

Active shootings represent a very small percentage of on-campus university violence.

But among those that do happen, there are patterns. And as law enforcement officials continue to investigate the Dec. 13, 2025, Brown University shooting, similarities can be seen with other active shooter cases on college campuses that scholar James Densley has studied. “They tend to happen inside a classroom, and there tends to be multiple victims,” Densley explains.

The Brown University tragedy, in which a shooter killed two students and injured nine more, marks the fourth deadly shooting at a U.S. university in 2025.

The Department of Education in Rhode Island, where Brown University is located, said on Dec. 16 that it is urging local elementary and secondary schools to review safety protocols.

Amy Lieberman, the education editor at The Conversation U.S., spoke with Densley about how schools have been given what he describes as an “impossible mandate” to try to prevent shootings.

What is the overall trajectory of school shootings over the past few years?

K-12 school shootings appear to be trending downward, at least in the past two years. But we actually saw the largest jumps in this type of violence in the three to five years leading up to 2024, which trends closely with the broader rise in homicide and violent crime we saw in the pandemic era.

In 2025, there have been 230 school shooting incidents in the U.S. – still a staggeringly high number. This compares with 336 school shootings in 2024, 352 in 2023, 308 in 2022, and 257 in 2021.

How this relates to an increase in schools trying to institute security measures to prevent shootings is an open question. But it’s true that many schools are experimenting with certain solutions, like cameras, drones, AI threat detection, weapons scanners, panic apps and facial recognition, even if there is only weak or emerging evidence about how well they work.

Schools are treated as the front line, because the larger, structural solutions are too difficult to confront. It is much easier to blame schools after a tragedy than to actually address firearm access, grievance pathways – meaning how a person becomes a school shooter – and the other societal problems that are creating these tragedies.

How have schools responded to the rise of school shootings in recent years?

Schools are being asked to solve a societal gun violence problem that they didn’t create and they cannot control. Even the best-run school cannot eliminate all risks when causes accumulate outside of their purview. These attacks are rare but catastrophic, and they create an impossible mandate for schools because when they occur, schools are told it reflects a failure in their preparation. Educators are expected to be teachers, social workers, threat assessors and first responders. It normalizes fear and shifts the responsibility downward.

There is a growing school safety industry that markets fear as a solvable, technical problem. It promises faster ways to detect weapons, for example, but the evidence base for those products is thin, proprietary or nonexistent. One example is an AI detection software that mistook a bag of Doritos for a gun, resulting in a large police response.

Schools are pressured to buy something from these companies to show they are doing something. But some of these systems create false positives, and, more importantly, they shift attention away from human relationships. Technology alone cannot resolve grievances, replace trust and create belonging, but most schools are focused on technology as a means of prevention.

How effective are other prevention systems schools have put in place?

If a school shooter is an outsider trying to attack the building, having a single point of entry, access control or multiple locks on doors creates time and space, which are essential for delaying an attacker until law enforcement can arrive, thus mitigating casualties.

But the evidence shows that nearly all school shooters are either current or former students at the school. They are very familiar with entry and exit points, and they are potentially already inside the building before the school can act on a potential threat of violence.

So, what happens if a school locks down, but you are actually locking the shooter in a room with their potential victims? What if students are forced to hide when it would be safer to run? What if you have a door that locks only from the inside and a student or staff member uses that room to bully or sexually assault another student? We’re building schools to protect against the rare events, but we are not mitigating the more common problems they face.

Students are being asked to practice preventing their own deaths in active shooter drills and learn in environments designed around worst-case scenarios. In general, interpersonal violence and spillover of community violence, like gang-related shootings, are the most common form of school shooting. Most shootings at schools occur in parking lots or at sports events, but we do very little to prepare for those types of scenarios.

Are there any benefits, then, to schools having certain non-tech safety measures in place, like making sure every person has an ID?

Of course, you don’t want strangers walking around in a school building. The fact that someone coming to the school has to get their ID scanned and wear a badge makes perfect sense, not just to prevent shootings but to also prevent theft and assaults and other risks.

The paradox is that school shooters tend to be children already affiliated with the school, and when someone walks in already firing, checkpoints and metal detectors are useless. Historically, several mass shootings in K-12 schools have started outside of the building then moved inside. The issue is not slipping past barriers but overwhelming them in seconds with irresistible force.

Absent policy change, what is the clearest way to prevent school shootings, according to current evidence?

Evidence shows that we often see signs of a crisis or withdrawal beforehand from school attackers. And that is why school-based behavioral threat assessment and management is so important. It is really about noticing changes in behavior and having the authority to intervene early. This is not about profiling people or relying on law enforcement alone. It is about having a structured, team-based process for identifying concerning behavior, assessing risk and coordinating appropriate supports – such as counseling – to prevent harm before it occurs. So often in these cases, people had a gut feeling that something was off with a particular student, but they didn’t know what to share or who to share it with.

For decades we’ve invested far more in responding to school shootings once they occur rather than in preventing them. You can lock doors and run drills, but no school can become a fortress.

Attackers leak warning signs in advance. Real prevention is about creating human systems that get upstream of this.

James Densley has received funding from the National Institute of Justice, the Joyce Foundation, and the Sandy Hook Promise Foundation.

​Politics + Society – The Conversation