Month: July 2026

Chatrie v. United States: The Warrant Rule for Location Data Just Got Simpler… and Harder

WRITTEN BY: Brian Tully

On June 29, 2026, the Supreme Court decided Chatrie v. United States. If you work cases that touch location data, this one impacts your work. The takeaway: when police obtain a person’s location data from a company like Google, that is a Fourth Amendment search, and you need a warrant. It does not matter the length of your request.  If it is location data held by a third party, get a warrant.

The Robbery

In 2019, a man robbed a credit union outside Richmond, Virginia. Witnesses and camera footage showed him approaching from a nearby church while appearing to talk on a cell phone. The man walked in and demanded $100,000 from the teller.  When the teller said she did not have access to that much cash, the man forced the manager at gun point to open the vault and left with $195,000. 

Investigators applied for a geofence warrant to Google, drawing a 150-meter circle around the scene and asking for the phones inside it around the time of the robbery.

The warrant used the three step Google process. Step one produced anonymized data for 19 phones inside the geofence. Step two narrowed the list to nine and pulled two hours of movement, inside and outside the circle. Step three narrowed to three and returned names. One of them was Okello Chatrie, whose data placed him at the scene about ten minutes before the robbery. Chatrie moved to suppress the evidence.

The lower courts split badly. The trial judge said the warrant violated the Fourth Amendment but admitted the evidence under the good faith exception. A Fourth Circuit panel said there was no search at all. The full Fourth Circuit deadlocked seven to seven on that question. The Supreme Court took one issue: was this a search?

What the Court Decided

Writing for the majority, Justice Kagan held that a person has a reasonable expectation of privacy in cell phone location data, and that accessing Location History is a search. The Court leaned on Carpenter v. United States, the 2018 CSLI case, and found that everything that made CSLI a search applies just as strongly, or more strongly, to Location History.  Having found a search, the Court sent the case back to the Fourth Circuit to decide whether this multi-step warrant actually satisfied probable cause and particularity at each step.

The specific geofence Google technique is largely gone already. In July 2025, Google moved Location History onto individual devices and now says it cannot respond to these warrants at all. So, the exact technique in Chatrie is off the table, but do not dismiss the importance of the decision.  The Court wrote broadly, and its reasoning reaches any location data held by any provider.

What This Means for Your Cases

1. Get a Search Warrant every time. Do not rely on the old arguments. Forget the 2703(d) orders.  Do not lean on a short time window. If you are seeking location data from a company, assume you need a warrant supported by probable cause.  For years, many state courts have required Search Warrants for any location data.  Now that rule is nationwide.
2. The reasoning travels. The Google technique is dead, but the logic covers Apple, Uber, Lyft, Snapchat, and whatever comes next. Do not treat a different provider as a warrant free workaround.
3. Remember Particularity. The dissent noted that when a request sweeps in private location data you did not need, the remedy is suppressing that piece, not the whole case. Narrow requests protect the rest of your evidence.  Be able to explain why you chose the geographical size of the request (in Chatrie, the 150-meter radius) and the time frame.   Avoid Constitutionally sensitive areas.  In the written and oral arguments, much was discussed that the radius of the Chatrie geofence warrant contained a church.  If your location request includes places of worship, abortion clinics, political organizations, or medical facilities, be able to explain why these locations are in the Search Warrant request.
4. Watch the horizon. The dissent warned that search history, purchase logs, and payment records could be next. The doctrine has not moved there yet, but if you are building around sensitive third party digital records, the safe course is a warrant.

Bottom Line

Chatrie makes the rule cleaner. When you want location data from a provider, get a warrant. The hard part moves to the paper. Write the affidavit and the warrant so that every step stands on its own probable cause and its own particularity.

Brian Tully recently retired from the Massachusetts State Police and is considered an SME regarding digital forensic evidence.

Algorithms Writing Reports

The Legal and Constitutional Stakes of AI Assisted Police Reports

WRITTEN BY:  Mike Schentrup

As police departments adopt generative artificial intelligence (AI) to draft incident reports from body-camera audio and field notes, the practice is colliding with state legislation, judicial precedent, and constitutional guarantees. While law enforcement tech companies market these tools as time saving innovations, legal advocates and civil liberties organizations warn that offloading narrative creation to software introduces severe risks to due process and the search for truth.

The Statutory Push for Transparency

State legislatures are increasingly taking action against opacity in AI-generated reports:

• Mandatory Disclosure & Audit Trails: Laws like California’s SB 524 and Utah’s SB 180 require explicit disclosure whenever AI assists in drafting a police report.
• Preservation Requirements: Agencies are obligated to retain original, unedited AI drafts alongside underlying audio files. This requirement ensures defense teams can trace how a narrative evolved from initial algorithmic output to the sworn document.
• Human-in-the-Loop Safeguards: Statutes mandate that officers review and individually sign off on report accuracy, attempting to preserve human accountability amidst automation.
• Local Jurisdictions Ban AI: There are several large municipal governments that have banned the use of AI to write police reports, including Seattle, Washington.

Judicial Scrutiny and Evidentiary Challenges

In courtroom proceedings, evidentiary rules and witness memory standards present steep hurdles for AI-assisted narratives:

• Memory “Contamination”: When an officer reads an AI summary generated from body-camera audio before drafting their final statement, the AI summary can overwrite or distort the officer’s actual subjective memory of the event. Under evidentiary rules governing refreshed recollection, defense attorneys argue that this process taints the witness’s independent recollection.
• Impeachment Evidence: Cross-examinations routinely exploit subtle discrepancies between an officer’s direct impressions, body-camera footage, and the text generated by AI.
• CJIS & Privilege Compliance: Federal compliance guidelines strictly prohibit the transmission of some unredacted police data to third-party cloud tools, raising data security and privilege waiver concerns.  This is especially true when using open source platforms, such as Chat GPT and Claude. This is not as much of an issue with proprietary closed systems, like Axon’s Draft One.

Constitutional Concerns

• Fifth and Fourteenth Amendment Due Process & Brady Obligations

• Brady v. Maryland requires prosecutors to disclose exculpatory or impeaching evidence. If an AI “hallucinates” details, glosses over an officer’s hesitation, or smooths over inconsistencies in a narrative, it obscures critical exculpatory material.
• The American Civil Liberties Union (ACLU) notes that algorithmic tools operate as opaque “black boxes,” hindering defendants from probing how a narrative was shaped or altered.
• Sixth Amendment Confrontation Clause
• The Confrontation Clause guarantees a defendant the right to cross-examine their accusers. When an incident report relies on a machine-synthesized narrative, the defense is left questioning a hybrid entity: an officer whose memory has been filtered through a proprietary algorithm.
• Fourteenth Amendment Equal Protection & Embedded Bias
• LLMs are trained on massive public datasets that contain historical societal and systemic racial biases. Is it possible these will be reflected in the AI assisted report narrative?
• Fourth Amendment Integrity
• Police reports often serve as the sole factual basis for probable cause determinations, arrest warrants, and search warrants. If automated drafting tools insert unverified framing or exaggerated details, warrants may be issued on inflated claims, undermining Fourth Amendment protections against unreasonable searches and seizures.

Ultimately, while automated tools offer speed, the combined weight of statutory mandates, evidentiary standards, and constitutional guarantees makes clear that efficiency cannot come at the expense of procedural fairness and human accountability.

Sources

1. https://statescoop.com/aclu-police-generative-ai-draft-reports/

2. https://admin.thehub.news/tag/aclu/feed/

3. https://statescoop.com/aclu-police-generative-ai-draft-reports/

4. https://admin.thehub.news/tag/aclu/feed/

Mike Schentrup is the owner/founder of Advanced Police Concepts and teaches report writing and supervisor report review across the United States.

Acting Corporal – A Leadership Dilemma

WRITTEN BY: Sgt Steve Sweeting

I love how police work, life, and philosophy often come together to teach a lesson. They can also combine in ways that reveal an old lesson in a new way.

This past month, I faced a leadership dilemma at work. It’s July, which means it is the busy season for us patrol cops. I have a team to lead through the busiest time of the year while my corporal is out for the month. So, when I need leadership and administrative backup the most, I find myself shorthanded. At the same time, it is promotion season, so a few of my officers are trying to carve out time to study and prepare for the upcoming test.

I was thinking about my lack of managerial assistance, the increased summer workload, and the fact that some of my key players were distracted by the upcoming promotions when a solution presented itself. Those of you who have been reading for a while have probably guessed correctly that I’m leading up to applying a lesson from Marcus Aurelius: “What inhibits action advances action. What stands in the way becomes the way.” Or, in popular terminology, “The obstacle is the way.” That is where I went with it, but that’s not the point of this particular newsletter.

I applied the old tactic of using problems against themselves and created a new job title on my team: “Acting Corporal.” It’s a completely fabricated job title, but it describes the role well. In short, I told each of my officers who would soon be testing for the job of corporal that they would take turns serving as the acting corporal. They would report directly to me, and everyone else would run things by them. All the small decisions and all the day-to-day issues were up to them. At several points during the day, I planned to meet with my acting corporal. They would review the paperwork and decide whether to approve or deny the reports, with my oversight, of course. I couldn’t delegate everything to them, but I was able to put enough on them that it felt real. They started feeling the weight of responsibility in a new way.

Check out Steve’s YouTube page @SteetStoicSteve

This completely made up job title worked better than I could have ever imagined. The acting corporals took it seriously and started trying to fill the role of corporal. They started having the hundreds of small interactions that leaders have on a daily basis. They started seeing how no two problems are identical, but usually they are all the same. They got to see first-hand how most of their coworkers were causing their own problems.

These lessons really crystallized during the report review portion of the night. They were harder on each other in terms of report writing and case management standards than I would ever be. They were quick to deny and slow to approve. Even though they were using my laptop and were logged in under my account, one officer looked at a lackluster report and said, “I’m not putting my name on this crap.” I had to laugh because it was actually my name going on the approval page, and the same officer had sent me many lackluster reports in the past.

The weight of responsibility and seeing the cases from a different perspective were changing how they perceived what was good enough and what was acceptable.

The real lesson came when I reviewed the cases and the deficiencies that my acting corporals were taking issue with. For each task that wasn’t done and each important thing that was overlooked or not documented, I asked them, “Why do you think the officer made this mistake?” I asked, “Do they not know?” I asked, “Do they lack experience?” or “Is this just an honest mistake?” Each and every one of them ultimately came to the same conclusion: The flaws in the cases were not caused by a lack of knowledge, resources, or experience. The officer simply didn’t want to do the task. They knew better but didn’t feel like doing it.

With each acting corporal, I discussed the phone calls, issues, and questions they had handled for their coworkers during the shifts. I asked whether there was any connection between the content of those conversations and what they were seeing in the reports. Each of them understood exactly what I was driving at. In many of those calls and conversations, the officers asking for advice already knew the answer. They knew what should be done, and they were usually calling or seeking advice because they wanted permission not to do the things they didn’t feel like doing.

It literally took only a day or two in a supervisory role for them to see the most common problem in police work and the primary obstacle to every good investigation and every good career: not doing the hard thing, especially the things you don’t feel like doing.

After each of them saw it for themselves and experienced it first-hand while the lesson was still fresh, I drove it home. This is a good leader’s job: protecting people from themselves. A good leader looks at the big picture, considers the long term, and guides people toward success. People are universally their own worst enemies and struggle to perform when they don’t feel like doing something. This is why parents make their children eat their vegetables. This is why coaches make their teams run laps in spring training. This is why we put the family on a spending budget. What we feel like doing is often the wrong answer. What we don’t feel like doing is usually what most needs to be done.

I’m happy that I invented a job title this month. I think the lessons learned did a lot of good for my aspiring leaders. I know that when they have their promotional interviews, they will have something insightful to say. I think I also did myself a favor by letting them learn that lesson. When they are out there working calls unsupervised, they will be more likely to recognize when they are avoiding necessary tasks simply because they don’t feel like doing them. Hopefully, that will mean fewer headaches for me in the future.

This idea holds true for working cases and writing reports, and it is also true in life. How much more successful would we all be, and how much closer to our goals would we all be, if the phrase “I don’t feel like it” didn’t exist? What if we learned to identify the thought, “I don’t want to,” as a massive green light and a directional arrow pointing us toward the things that we absolutely should do? I think we would all live with fewer consequences and regrets. I know we would all be more successful if we did.

Until next week, BOLO for the dangers of “I don’t want to.”

Amor Fati