Ross v. United States

• Court: District of Columbia Court of Appeals
• Decided: February 20, 2025
• Docket: 23-CM-1067
• Status: Published

Opinion

Notice: This opinion is subject to formal revision before publication in the Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the Court of any formal errors so that corrections may be made before the bound volumes go to press.

DISTRICT OF COLUMBIA COURT OF APPEALS

No. 23-CM-1067

NIYA ROSS, APPELLANT,

v.

UNITED STATES, APPELLEE.

Appeal from the Superior Court of the District of Columbia (2023-CMD-006252)

(Hon. Hiram E. Puig-Lugo, Trial Judge)

(Submitted December 5, 2024 Decided February 20, 2025)

Sweta Patel was on the brief for appellant.

Matthew M. Graves, United States Attorney at the time the brief was filed, and Chrisellen R. Kolb, Nicholas P. Coleman, Jasmine Dohemann, Nickolas Reck, and Kevin Birney, Assistant United States Attorneys, were on the brief for appellee.

Before DEAHL, HOWARD, and SHANKER, Associate Judges.

Opinion for the court by Associate Judge SHANKER.

Concurring opinion by Associate Judge HOWARD at page 20.

Dissenting opinion by Associate Judge DEAHL at page 27.

SHANKER, Associate Judge: Leaving a dog in a car on a hot day is always

inadvisable. But is it always criminal? A statute tells us when it is, and, as with all 2

crimes, the government must prove the elements of the statute beyond a reasonable

doubt with evidence, not with appeals to common knowledge, common sense, or

common understandings of right and wrong. Res ipsa loquitur is not a doctrine in

the criminal law.

In this case, on a hot day in early September 2023, appellant Niya Ross left

her dog Cinnamon in her car parked in the shade of a tree, with the windows lowered

a few inches, for over an hour. Following a bench trial, Ms. Ross was convicted of

animal cruelty in violation of D.C. Code §§ 22-1001, 22-1002. At trial, the

government presented little if any evidence regarding the temperature within the car,

the weather conditions that Ms. Ross’s dog could comfortably tolerate, or symptoms

of heat-related distress manifested by the dog.

On appeal, Ms. Ross argues that the evidence was insufficient to support her

conviction. The government responds that it is “common knowledge” that a dog

would suffer in the conditions present in Ms. Ross’s car that day. Because we do

not believe common knowledge suffices to fill critical gaps in the government’s

case, we reverse Ms. Ross’s conviction and remand with instructions to enter a

judgment of acquittal. 3

I. Factual and Procedural Background

Viewing the evidence in the light most favorable to the verdict, the trial court

could have found the following facts. On the hottest September fourth then on record

in the District—reaching a high of ninety-eight degrees Fahrenheit1—Zachary

Vasile was walking down N Street, NW, toward a supermarket when he heard

“loud,” “incessant dog barking.” Turning toward the sound, Mr. Vasile noticed a

dog—a “doodle” with “very dense sort of curled hair”—alone inside a car with the

windows cracked open approximately three to five inches. As Mr. Vasile

approached the car, the dog noticed him and stopped barking.

Because of the “very hot” day and the “very loud,” “constant[ ]” nature of the

dog’s barking, Mr. Vasile “realized this was a potentially dangerous situation for the

dog.” He shouted for the dog’s owner and attempted to open the door of the vehicle,

both to no avail. After about “five to six minutes” of shouting, Mr. Vasile called

311 and explained the situation.

The 311 dispatcher sent “fire, animal control, and police” to Mr. Vasile’s

location, putting out an “all call” for a “level [one] emergency,” defined as an animal

1 The trial court took judicial notice of “a printout from the National Weather Service, including temperature for Washington, D.C., on September the 4th, 2023.” Around the time of the incident, the dash thermometer of a responding police cruiser read approximately ninety-seven degrees Fahrenheit. 4

“actively . . . in distress with a risk of death or great bodily injury.” Firefighters (the

first on the scene) initially “looked around, trying to get a view of the dog.” They

then forced a window further open, unlocked the car, and let the dog out. Shortly

thereafter, Aristides Torres—an animal control officer—arrived. Officer Torres

scanned the dog’s microchip, used the scan results to identify the owner’s phone

number, and called the number repeatedly for ten to fifteen minutes. No one

answered.

During his efforts, Officer Torres placed the dog inside his van. Once police

officers arrived, Officer Torres opened the van to show them the dog. At that time,

bodycam footage showed the dog’s tongue sticking out, suggesting that the dog was

panting.

Mr. Vasile left the scene after approximately forty minutes, during which time

the dog’s owner did not return to the car. Police officers, however, remained on the

scene until Ms. Ross arrived at 6:11 p.m.—around an hour after responders were

dispatched in response to Mr. Vasile’s call. Officers approached Ms. Ross as she

was unlocking the car, asking, “Where’s your dog?” After Ms. Ross confirmed the

dog belonged to her, officers placed her under arrest. Body-worn camera footage

taken during the arrest shows Ms. Ross’s car parked in the shade of an adjacent tree. 5

The government charged Ms. Ross with one count of cruelty to animals under

D.C. Code §§ 22-1001, 22-1002. At a bench trial, Ms. Ross contended that (1) the

government failed to prove—particularly in the absence of expert testimony—that

the dog suffered as a result of being left in the car and (2) she lacked the requisite

intent for a conviction, relying in part on the testimony of a character witness. In

support, she pointed out that Mr. Vasile did not testify regarding any symptoms of

heat-related distress on the part of the dog. By contrast, Officer Torres affirmatively

testified that when he arrived at the scene, the dog “didn’t look like it . . . was in

distress.”

In response, the government referenced evidence showing that the dog was

barking incessantly, that the dispatcher treated Mr. Vasile’s report as a level one

emergency, and that Officer Torres, who had “responded to a dog locked in a [hot]

car” more “than [fifty] times,” testified that if the dog had been left in the car longer,

“it could definitely have been fatal.” The government also noted that Ms. Ross—a

former pediatric nurse—testified that she had never left a child in a hot car for an

hour because “that’s just not right.”

The trial court found Ms. Ross guilty. It explained that “the fact that Ms. Ross

left not one, not two, but four windows rolled down a distance of about three to five

inches shows that she knew that there was a plain and strong likelihood that harm 6

might resolve to Cinnamon because of the heat.” It also found that there was no

justification for Ms. Ross’s actions, as she did not remain within earshot of

Cinnamon and left her alone for “roughly an hour and [fifteen] minutes,” not

including “how long [Cinnamon] had been sitting in the car before Mr. Vasile’s

arrival.” Finally, the trial court stated, “[Y]ou don’t need expert testimony to

conclude that leaving a dog in a hot car for at least an hour and [twenty] minutes is

failing to provide the dog with protection and shelter from the weather.”

This appeal followed.

II. Analysis

Ms. Ross argues that the evidence presented at trial was insufficient to sustain

her conviction, presenting two arguments to support this conclusion. First, she

contends that, particularly in the absence of expert testimony, the government failed

to demonstrate that “the circumstances in which [Cinnamon] was found” caused

Cinnamon to suffer. Second, she attacks what in her view is the absence of evidence

that she possessed the required mens rea—general intent with malice.

After discussing our standard of review and the elements of the District’s

animal cruelty statute, we turn to Ms. Ross’s arguments. As we agree with

Ms. Ross’s first argument, we reverse without reaching her second. 7

A. Standard of Review

We review a claim of insufficient evidence de novo, considering “all the

evidence in the light most favorable to the verdict[ and] according deference to the

factfinder to weigh the evidence, determine credibility, and draw justifiable

inferences of fact.” Wicks v. United States, 226 A.3d 743, 746-47 (D.C. 2020)

(internal quotation marks omitted). It follows that, where bench trials are concerned,

we “are deferential to the prerogatives and advantages of the trial judge” and “will

not disturb the trial judge’s factual findings unless we can conclude they were plainly

wrong or without evidence to support them.” Augustin v. United States, 240 A.3d

816, 824 (D.C. 2020). We affirm the trial court’s judgment if “any rational trier of

fact could have found the essential elements of the crime beyond a reasonable

doubt.” Rivas v. United States, 783 A.2d 125, 134 (D.C. 2001) (en banc) (internal

quotation marks omitted).

But this review “means more than that there must be some relevant evidence

in the record in support of each essential element of the charged offense”; “[s]light

evidence is not sufficient evidence.” Id. at 134. Indeed, evidence is not sufficient

unless it “eliminate[s] the possibility that the [factfinder’s] verdict was based on

surmise or conjecture.” See Long v. United States, 156 A.3d 698, 713 (D.C. 2017)

(internal quotation marks omitted). 8

B. Elements of Animal Cruelty

D.C. Code § 22-1001 provides, in relevant part, that anyone with “charge or

custody of any animal” who “unnecessarily fails to provide the same with proper

food, drink, air, light, space, veterinary care, shelter, or protection from the weather”

is guilty of a misdemeanor. And, although § 22-1001 uses the word “knowingly,”

we have interpreted it to require a mens rea of “general intent with malice.”

Dauphine v. United States, 73 A.3d 1029, 1032-33 (D.C. 2013). So, to sustain

Ms. Ross’s conviction for animal cruelty, we must find sufficient evidence that

(1) she failed to provide Cinnamon with proper protection from the weather, (2) she

“had no justification, excuse or recognized mitigation” for her actions, and (3) she

“was at least aware of the plain and strong likelihood” that harm may result,

Dauphine, 73 A.3d at 1032-33 (internal quotation marks omitted).

Where prong one of the above test is concerned, generalized evidence is not

sufficient evidence. Instead, evidence of improper protection from the weather must

be particularized to both the dog and situation at hand. See Jordan v. United States,

269 A.2d 848, 849 (D.C. 1970) (“In the absence of testimony . . . that the shelter or

protection from the weather supplied this dog on this occasion would cause the dog

to suffer, the evidence was insufficient to sustain the conviction.” (emphasis added

and footnote omitted)). 9

C. Proper Protection from the Weather

With the preliminaries settled, we turn to Ms. Ross’s contention that the

evidence was insufficient to support, beyond a reasonable doubt, the conclusion that

she failed to properly protect Cinnamon from the weather. We agree with Ms. Ross

that the government’s case was insufficient as a matter of law. Although the record

supports the conclusion that some dogs, if left in a hot car for some undefined period

of time, would suffer, no reasonable finder of fact could conclude beyond a

reasonable doubt that Cinnamon suffered or would have suffered on September 4,

2023.

We note at the outset that the evidence presented at trial does not establish

two critical facts: (1) the actual temperature inside the car and (2) the presence of

symptoms of heat-related distress in Cinnamon.

The government does not contend on appeal that it proved either of the above

two facts. Instead, it suggests that it could rely on the factfinder to infer those facts

by applying its common sense to other evidence in the record. According to the

government, it was enough for it to prove only that (1) the temperature outside of

the car was approximately ninety-eight degrees Fahrenheit and (2) Cinnamon would

have (had she not been released) remained within the car for approximately one hour

and twenty minutes. 10

We disagree. To be sure, a factfinder may, as a general matter, use their

common sense and everyday experience to draw “reasonable inferences from the

evidence presented in a trial.” Long, 156 A.3d at 714. But where these “common

sense” inferences are grounded in the assumed nonexistence of mitigating facts, they

will often “cross[ ] the line . . . into the prohibited territory of surmise and

conjecture.” See id. at 714-15. Reasonable inferences must be drawn from, and

common sense applied with respect to, evidence; inferences and common sense

cannot serve as substitutes for evidence. See id. at 714. Cf. Galloway v. United

States, 319 U.S. 372, 387 (1943).

The facts in Long amply illustrate this principle. There, we held that the

evidence was insufficient for a jury to conclude beyond a reasonable doubt that a

stolen, ten-year-old Dodge Intrepid was worth at least $1,000 absent evidence of the

“cost of the Intrepid when new,” “the price [the complainant] paid for it at” auction,

or the car’s “mileage, maintenance history, or Bluebook value at the time of the

robberies.” Long, 156 A.3d at 714. This was so even though the evidence

established that (1) the car was “pretty clean,” (2) the robbers were able to drive the

car at significant speed during a police chase, (3) the car was operable throughout

the more-than-two-year period between when the complainant recovered it and the

robbers’ trial and (4) the body of the car was in “decent condition, with no significant

dents or scratches beyond what one would ordinarily expect to see on a ten-year-old 11

car in a busy urban area.” Id. In short, where the jury could not rule out the existence

of facts that would reduce the value of the car, their “common sense” valuation could

not make up the difference.

One might think that a functional car would obviously have been worth more

than $1,000.2 That instinct, however, is not enough; Long teaches that a hunch as to

whether a fact is obvious—even one supported by common sense and everyday

experience—does not suffice where mitigating circumstances could reasonably have

2 We, like the dissent, harbor doubts about whether ChatGPT is “a good proxy for what is, and what isn’t, common knowledge.” Post at 37 n.4. But to the extent the dissent relies on ChatGPT’s degree of certainty as to whether leaving a dog in a hot car is harmful, we note that ChatGPT expresses similar certainty in its response to the query, “How much would a pretty clean, operable 2002 Dodge Intrepid with no significant dents or scratches have been worth in 2012?”: In 2012, a 2002 Dodge Intrepid in pretty clean, operable condition with no significant dents or scratches would have likely been worth between $3,000 and $5,000. The exact price would have depended on factors such as mileage, location, and the specific condition of the car (interior, mechanical health, etc.). For instance, a lower- mileage model with a well-maintained engine could have been closer to the higher end of that range, while a higher- mileage model might have been valued closer to the lower end. Generally, as a mid-size sedan from 2002, it would not have been fetching a premium price by 2012, but it still had decent value as an affordable used car. Note that $3,000 (ChatGPT’s minimum price) is significantly above the $1,000 threshold at issue in Long. ChatGPT’s analysis, therefore, does not correspond to at least one of our sufficiency decisions. 12

been in play. See 156 A.3d at 714-15; see also Mejia-Cortez v. United States, 256

A.3d 210, 214 (D.C. 2021) (explaining that “due process generally requires actual

proof beyond a reasonable doubt” of each “critical fact,” “even where the fact in

question may seem incontrovertible”). With this conclusion in mind, we now turn

to the case before us.

Where Cinnamon is concerned, a reasonable factfinder would have found that

mitigating circumstances could have reasonably been in play. We reach this

conclusion for two reasons: (1) the specific conditions in which Cinnamon was

found suggest that the temperature within the car may have been cooler than the

unshaded surrounding area and (2) testimony established that Cinnamon manifested

no symptoms of heat-related distress despite spending approximately forty minutes

inside the car before she was let out.

We begin with the conditions in which Cinnamon was found. We, like our

dissenting colleague, would balk at entering a closed car, parked in full sun, on a

summer’s day. But we seek out shaded parking and leave windows cracked open to

avoid just such an unpleasant circumstance. Here, no witness testified that the car

was exposed to direct sunlight, the body-worn camera footage admitted by the

government seems to show Ms. Ross’s car parked in the shade, and Ms. Ross left all 13

four of the car’s windows open between three and five inches. 3 It seems to us,

therefore, that the interior of the car could well have been cooler than the car’s

unshaded surroundings. And this possibility, in our view, creates a reasonable doubt

as to whether Cinnamon was enduring harsh conditions inside the car.

Our conclusion in this regard is bolstered by Officer Torres’s testimony that,

despite having spent approximately forty minutes inside the car, Cinnamon “looked

fine,” did not appear “in distress,” and seemed in good health. If the interior of the

car were in fact hot enough to cause a dog like Cinnamon to suffer, we would expect

Cinnamon to have manifested at least some symptoms of exposure to heat. Indeed,

in each of the out-of-jurisdiction cases in which an appellate court has found the

evidence sufficient to sustain an animal-cruelty conviction for leaving a dog in a hot

car, the dog presented just such symptoms. See State v. Butler, 293 A.3d 191, 194

(N.H. 2022) (explaining that the confined dog post-rescue registered an internal

temperature of “over 105 degrees” and required “24-hour-care” in a “veterinary

hospital”); State v. Washburn, 325 A.3d 136, 143 (Vt. 2024) (referencing the

confined dog’s “panting,” “swollen” tongue, and diagnosis with the “early stages of

3 The trial court viewed the fact that the windows were lowered as evidence of Ms. Ross’s knowledge that leaving a dog in a car on a hot day is dangerous, but the fact certainly undermines any assumption that the temperature inside the car soared as it would in a completely closed car. 14

heat exhaustion”); City of Beachwood v. Pearl, 111 N.E.3d 620, 636 (Ohio Ct. App.

2018) (relying on testimony that the confined dogs were “panting profusely with

very dry, hot tongues,” were “beginning to become distressed,” and were “showing

signs of heatstroke or dehydration”).4 The absence of such symptoms here cements

our conclusion that a reasonable factfinder must have had a reasonable doubt in this

case.5

Before turning to the response of the government and the dissent, we pause to

underscore why Long and Mejia-Cortez encourage caution where a factfinder relies

on “common sense” assumptions: what we believe to be true may not be so. Take

the inference that shade and ventilation could not have kept the temperature of

Ms. Ross’s car within a tolerable range. A Department of Energy website suggests

that “air temperatures directly under trees can be as much as 25°F cooler than air

The dissent points out that in these three cases, our sister courts found the 4

evidence sufficient despite the dogs having been “left in cars for both briefer periods and on cooler days.” Post at 40. This critique misses the mark—we are not suggesting that no dog left in a car on a ninety-eight degree day for an hour and twenty minutes will suffer. Instead, we contend that because, despite being left within a car for a longer period and on a hotter day than those other dogs, Cinnamon did not manifest their symptoms, we see a reasonable probability that Cinnamon was differently situated from those other dogs. Put differently, the absence of symptoms of heat exhaustion in Cinnamon reinforces the conclusion that mitigating circumstances (e.g., shade, ventilation) were at play in her case. Because this holding alone suffices to reverse Ms. Ross’s conviction, we 5

need not address Ms. Ross’s contention that the evidence was insufficient to prove that she acted with malice. 15

temperatures above nearby blacktop.” Landscaping for Shade, U.S. Dep’t Energy,

https://www.energy.gov/energysaver/landscaping-shade; https://perma.cc/DB76-

HYJ6. If the Department of Energy’s estimate is correct, the temperature inside of

the shaded car could have been seventy-three degrees Fahrenheit—twenty-five

degrees cooler than the temperature recorded by the police car. Or, to give another

example, the government suggests that Cinnamon’s fur acted “as an additional layer

that retains heat,” but the American Kennel Club explains that many dogs with

apparently thick fur have double coats that, by creating an insulating effect, in fact

help keep them “cool in hot weather.” Harriet Meyers, Is It OK to Shave Your Dog’s

Coat in Summer, American Kennel Club, https://www.akc.org/expert-

advice/health/is-it-ok-to-shave-your-dog/; https://perma.cc/PVH5-B477.

Cinnamon’s fur thus might cut against the government.

We express no opinion on the accuracy of the facts asserted by the above

sources; adversarial testing at trial could well have debunked them. We cite them

only to support the following point. When a factfinder engages in “common sense”

adjudication in place of adjudication grounded in evidence, it creates a greater risk 16

that its judgment will rest on factual error—the exact sort of risk that the “beyond a

reasonable doubt standard” was designed to reduce. See Rivas, 783 A.2d at 133.6

The government and the dissent respond to our conclusion that the evidence

was insufficient in two ways.

First, they dispute our conclusion that Cinnamon manifested no symptoms of

heat-related distress. For its part, the government suggests that Cinnamon “showed

outward signs of distress given her incessant barking.” But the government and

Ms. Ross offer equally plausible, competing explanations for Cinnamon’s barking—

the government reads it as a cry for relief from the heat, while Ms. Ross suggests it

stemmed from separation anxiety. As neither we nor the trial court speak dog—and

the government offered no witness who purported to do so—we are not convinced

6 The dissent recognizes that what a factfinder considers to be common knowledge may prove incorrect. We cannot agree, however, with the dissent’s implied solution—that a factfinder need only reexamine its assumptions when faced with evidence that “actually undermines or contradicts [them].” See post at 37. That contention runs headlong into the proposition, recognized by the trial court, that a criminal defendant is “not required to prove anything.” See also Porter v. United States, 826 A.2d 398, 407 (D.C. 2003) (“[E]very defendant in a criminal trial has a right not to testify or not to produce any evidence, and the burden of proving guilt rests with the government.”). We thus find the dissent’s reliance on Ms. Ross’s failure “to offer any contradictory evidence” ill-taken. Post at 30. 17

that a reasonable factfinder, applying the beyond-a-reasonable-doubt standard, could

adopt the government’s interpretation to the exclusion of Ms. Ross’s. 7

The dissent, rather than rely on Cinnamon’s barking, cites body-worn camera

footage (taken while Cinnamon was inside Officer Torres’s van) that it contends

shows Cinnamon “panting heavily.” Post at 41. It also notes that Officer Torres

identified excessive panting as a sign of heat-related distress. We, however, think

that Officer Torres is the better judge of what constitutes excessive panting. And

Officer Torres reported no signs of distress when he examined Cinnamon. We do

not believe that our obligation to view the evidence in the light most favorable to the

government requires us to supplant a diagnosis made by an animal control officer

present at the scene with our own “common sense.”

Second, the government and the dissent contend that Officer Torres’s

testimony filled any record gaps regarding the conditions within Ms. Ross’s car. So,

7 At the risk of sounding like a broken record, the government’s “common sense” inference regarding the cause of Cinnamon’s barking again appears to rest on a shaky foundation. The Royal Society for the Prevention of Cruelty to Animals does not list barking as a symptom of heatstroke in dogs. Dog Heatstroke Treatment, RSCPA, https://www.rspca.org.uk/adviceandwelfare/pets/dogs/health/h eatstroke; https://perma.cc/9SUR-TZUW. By contrast, the American Kennel Club lists “barking” as indicative of separation anxiety. Randa Kriss, Anxiety in Dogs: Signs, Symptoms, Treatment, American Kennel Club, https://www.akc.org/expert- advice/health/treating-dog-anxiety/; https://perma.cc/6KBU-VJ7M. To reiterate, however, we do not rely on this extra-record evidence for purposes of resolving this appeal. 18

let’s look at that testimony. Officer Torres testified that “dogs locked in cars can be

fatal, given the time frame of how long they’ve been in a car.” He explained that

the temperature inside a car, depending on its time of exposure, can be “double” the

exterior temperature. Finally, he stated that if Cinnamon had been left in the car for

longer, “it could definitely have been fatal.”

That’s a lot of caveats. We do agree, as a matter of common sense, with the

thrust of Officer Torres’s testimony. Yes, the interior of a car can, depending on the

car’s degree of exposure to the elements, become hotter than the exterior

temperature. And yes, locking a dog in a car can be fatal, depending on how long

the dog remains in the car and how hot the car’s interior becomes. Those

propositions are but truisms. What is missing from Officer Torres’s testimony is

specifics. For how much longer did Cinnamon need to be left in the car? At what

interior temperature does sustained exposure pose a risk of health consequences?

Officer Torres did not say. And even if he had said, he had no way of knowing

whether those thresholds were reached in Cinnamon’s case. How could he? After

all, he (1) “didn’t inspect the vehicle” and (2) thought Cinnamon “looked fine.”

At bottom, Officer Torres’s testimony rests on similar assumptions to those

we rejected above. To us, therefore, his testimony cannot fill the gaps we have 19

identified. 8 Accordingly, we hold fast to our conclusion that the evidence presented

in this case was insufficient.

By reaching this holding, we in no way intend to question the actions taken

by Mr. Vasile, the firefighters, Officer Torres, and the responding police officers.

To the contrary, we commend them for their heroic efforts; where a situation could

be dangerous for an animal like Cinnamon, we hope all would respond just as

Mr. Vasile did. And let us repeat: leaving a dog in a car on a hot day is at a minimum

ill-advised. The question here, though, is whether it was criminal. And where

criminal sanctions are concerned, “could be” is not enough. We thus reverse

Ms. Ross’s conviction.

8 We do not think we are being too stingy with respect to Officer Torres’s testimony. After all, we have not created the requirement that the government’s showing be particularized to the circumstances involved in the alleged cruelty out of whole cloth—that requirement flows from our statements in Jordan. See 269 A.2d at 849 (noting that the factfinder’s determination of guilt must focus on “the shelter or protection from the weather supplied this dog on this occasion” (emphasis added)). 20

III. Conclusion

For the foregoing reasons, we reverse and remand with instructions to enter a

judgment of acquittal.

So ordered

HOWARD, Associate Judge, concurring: I join the majority opinion and write

separately only to address a point of significant interest, raised by the writings of my

colleagues, which together represent the first published discussion involving the use

of AI tools in decision making at this court. Following in the recent footsteps of

Judge Newsom from the U.S. Court of Appeals for the Eleventh Circuit, the dissent

makes use of the ChatGPT large language model artificial intelligence tool by

OpenAI, and the majority opinion responds in-kind to contrast the effort. See post

at 37-39 & nn.4-5; ante at 11 n.2. Deferring discussion to an opinion with

precedential force, the dissent points to Judge Newsom’s thoughtful discussion in

the Snell case. See Snell v. United Specialty Ins. Co., 102 F.4th 1208, 1221-35 (11th

Cir. 2024) (Newsom, J., concurring). I write, since we have broached the topic, to

highlight a few brief points not addressed in Judge Newsom’s insightful

concurrence, that I have considered personally and as part of our D.C. Courts AI

Task Force, which I find important considerations in judicial and court use of AI

tools. 21

To be clear, I cast no aspersion on the use of AI by my colleagues. I find it

interesting. AI tools are proliferating and we ignore them at our own peril. Not only

for the concerning capabilities they now give parties with ill intent,1 but for the great

utility such tools could potentially provide in easing the strain on our increasingly

overburdened courts.2

1 Such as the ever-increasing ability to falsify evidence, like images, audio, and video, and sow disinformation in and outside of the courtroom. A prominent example is the memorable “deep fake” image, generated by a construction worker playing with an AI tool, of Pope Francis in a particularly stylish puffer coat, which appeared indistinguishable from reality and took the internet by storm. See Drake Bennett, AI Deep Fake of the Pope’s Puffy Coat Shows the Power of the Human Mind, Bloomberg (April 6, 2023, 7:00 AM), https://www.bloomberg.com/news/newsletters/2023-04-06/pope-francis-white- puffer-coat-ai-image-sparks-deep-fake-concerns; https://perma.cc/3D6V-ZCFW. A cogent description and discussion of these concerns can be found in the interim guidance from the National Center for State Courts. AI Rapid Response Team, Nat’l Ctr. for State Cts., AI and the Courts: Digital Evidence and Deepfakes in the Age of AI (June 2024), https://www.ncsc.org/__data/assets/pdf_file/0019/101683/ncsc-ai-rrt-deepfakes- june-2024.pdf; https://perma.cc/86VE-3WGM. 2 The D.C. Courts are such a system. It has been well covered in the news for several years that we face historic vacancies. The D.C. Superior Court has in the recent past, and may soon again, faced double digit vacancies. This court has been without a full complement of judges for more than a decade. Our current two-judge vacancy represents nearly one-third of our court. Compounding this issue is an ever- dwindling number of senior judges—former active judges who commit to continued public service—seasoned veterans who are valuable resources in accomplishing the work of this court. And, structurally, this court faces the challenge of having the same nine judgeships over the court’s approximate fifty-five-year existence, while the courts and agencies we review have grown to numbers of judges and administrative adjudicators regularly exceeding 100 each. Compare D.C. Code 22

AI tools are more than a gimmick; they are coming to courts in various ways,

and judges will have to develop competency in this technology, even if the judge

wishes to avoid using it. Courts, however, must and are approaching the use of such

technology cautiously. Specific use cases are being considered and we must always

keep in mind the limits of different AI tools in how and when we use them,

particularly with regard to security, privacy, reliability, and bias, to ensure ethical

use.

Broadly, an AI system can be susceptible to bias at multiple points in its

execution.3 Model Code of Judicial Conduct Rules 2.2 and 2.3, dealing with

impartiality and fairness and bias, prejudice, and harassment, are potentially

implicated in reliance on a system infected with bias. Ignorance of the technology

seems like little defense in consideration of the duty of competence in Rule 2.5.

Other issues abound, but security and confidentiality of court information are

particular concerns. Accordingly, before using an AI tool a judicial officer or staff

§ 11-702 (1973 Ed.), with D.C. Code § 11-702 (composition of D.C. Court of Appeals); see D.C. Code § 11-903 (composition of D.C. Superior Court); D.C. Superior Court Judges, D.C. Cts., https://www.dccourts.gov/superior-court/judges; https://perma.cc/Q6GP-VMGL; D.C. Public Salary Information, D.C. Dep’t Hum. Res., https://dchr.dc.gov/public-employee-salary-information; https://perma.cc/L8Q6-KLJA (listing all D.C. employees). 3 IBM offers a relatively short, simplified breakdown of AI bias. James Holdsworth, What is AI Bias?, IBM (Dec. 22, 2023), https://www.ibm.com/think/topics/ai-bias; https://perma.cc/HL7S-YJW7. 23

member should understand, among many other things, what data the AI tool collects

and what the tool does with their data.

The quote has many attributions that “if it is free, you are the product.” Many

AI tools benefit from what we feed into them, documents, prompts, etc., virtually

every aspect of our interaction trains and hones such tools. That is part of the early-

mover advantage of ChatGPT in particular, which blew away previous records to

reach one million users in five days 4—and 100 million within two months of going

live.5 As of January 30, 2025, it was estimated to have approximately 300 million

weekly users. 6 It is hard to imagine a company that could afford to pay that many

people to test and develop their model. However, such a system raises serious

practical and ethical issues for a court. Security is a preeminent concern. I briefly

look at a few hypotheticals in the context of this court to illustrate.

4 Bernard Marr, A Short History of ChatGPT: How We Got to Where We Are Today, Forbes, (May 19, 2023, 1:14 AM), https://www.forbes.com/sites/bernardmarr/2023/05/19/a-short-history-of-chatgpt- how-we-got-to-where-we-are-today/; https://perma.cc/5LCV-8TNH. 5 Krystal Hu, ChatGPT Sets Record for Fastest-Growing User Base – Analyst Note, Reuters (February 2, 2023, 10:33 AM), https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user- base-analyst-note-2023-02-01/; https://perma.cc/425D-HGQU. 6 ChatGPT / OpenAI Statistics: How Many People Use ChatGPT?, Backlinko, https://backlinko.com/chatgpt-stats; https://perma.cc/3QY3-ZN59 (last updated January 30, 2025). 24

First, take the use case of a judge utilizing an AI tool to summarize briefs filed

with the court well in advance of oral argument—a practice, along with summarizing

voluminous records, that some AI tools appear to be quite adept at. It is the practice

of this court to announce the members of a particular panel of judges the week before

an oral argument. Should a judge be using an AI tool that trains on the data they

submitted, they have now surrendered data which includes—at bare minimum—the

submitted data, i.e. the briefs of the parties, and potentially personally identifying

data, i.e. a username, IP address, and email address. Data which, reviewed together,

could expose the judge’s involvement on the panel to individuals and systems with

access to that data before that information is public.

Next, fast-forward past argument and assume our hypothetical technophile

jurist decides they will have the AI tool aid them in the preparation of a decision. AI

tools offer many potential use cases here. For one, perhaps with careful prompting,

detailing the types of facts or story that is desired, the AI tool could be used to pull

from the record and produce a first draft of the factual rendition section of the

decision. It could develop an initial statement of the standard of review and

controlling law. In varying degrees of quality, depending on the tool and inputs, it

could formulate a first take at some analysis. However, again, should the AI tool be

training itself on the data, someone with access to the data would have access to

judicial deliberative information and potentially personally identifying login/user 25

information that could identify the judge as well. Of even more concern, as the data

trains the tool, another user could stumble upon it or some aspects of it regurgitated

by the AI tool. Even if the odds are miniscule, confidential judicial deliberative

information has potentially leaked out ahead of a decision in this scenario.

Consider further the scenario that any of the material used in either prior

hypothetical contained sensitive information that would otherwise be subject to

redaction, i.e. social security numbers, account numbers, minor’s names, etc. If

unredacted briefs or records were loaded into the AI tool, it would be an instant

failure of the court’s duty to protect such information. Three hundred million users,

in the scenario of ChatGPT, described above, would potentially have access.

I pause briefly here to note that such concern does not appear to arise from the

use of AI in this decision. The dissent’s generalized hypothetical questioning,

without more, does not strike me as remotely unique to this case in a way that could

even inadvertently expose deliberative information. The majority’s use of ChatGPT

provides comparison by prompting the tool against the facts of a previous case for

analysis. It strikes me that the thoughtful use employed by both of my colleagues

are good examples of judicial AI tool use for many reasons—including the

consideration of the relative value of the results—but especially because it is clear

that this was no delegation of decision-making, but instead the use of a tool to aid 26

the judicial mind in carefully considering the problems of the case more deeply.

Interesting indeed.

The previous examples that I described as potential improper use of an AI

tool, however, could be accomplished with the use of an AI tool with robust security

and privacy protections. Even more exciting, AI companies have begun to announce

the release of government oriented tools which promise to provide such protections

and allow for such potential use cases.7

As state courts across the country cautiously consider these issues, the

National Conference of State Courts has taken a lead in coordinating efforts. It has

put together an AI Rapid Response Team and created a policy consortium,

constantly updating resources. See Thomson Reuters Institute/NCSC AI Policy

Consortium, Nat’l Ctr. for State Cts., https://www.ncsc.org/consulting-and-

research/areas-of-expertise/technology/tri-ncsc-ai-policy-consortium;

https://perma.cc/2Z66-GXB4. And the D.C. Courts have not stood idly by, creating

our D.C. Courts AI Task Force and partnering with the National Conference of State

7 A particularly interesting discussion of the most recent announcements with thoughtful comments on the potential for courts—complete with what appears to be an AI generated image of judge with a holographic heads-up-display—comes from Judge Scott Schlegel of Louisiana. See The Wait May Be Over: Government AI Products Could Give Courts the Green Light, [sch]Legal Tech, (February 5, 2025), https://judgeschlegel.substack.com/p/the-wait-may-be-over-government-ai; https://perma.cc/5KY5-7TD4. 27

Courts. See District of Columbia Courts Administrative Order: Artificial

Intelligence Task Force of the District of Columbia Courts, D.C. Cts. (March 5,

2024), https://www.dccourts.gov/sites/default/files/Artificial_Intelligence_Task_F

orce_Administrative_Order_Final.pdf; https://perma.cc/JCU7-V9FW. As the use of

AI begins to appear at the D.C. Courts, litigants and the citizens of the District can

be assured that cautious and proactive thought is being directed by our judges and

D.C. Courts team members, toward the beneficial, secure, and safe use of AI

technology.

DEAHL, Associate Judge, dissenting: Niya Ross left her dog alone in a parked

car in ninety-eight degree heat for an hour and twenty minutes. Ross parked her car

facing directly into the sun, with minimal shade coming off a nearby tree, and left

during the hottest portion of the hottest September 4 on record in the District’s

history—the late afternoon (from about 4:50 to 6:10 p.m.). Those are probably all

the facts necessary for most people to conclude beyond a reasonable doubt that

Ross’s actions created a “plain and strong likelihood” that her dog would be harmed.

See Dauphine v. United States, 73 A.3d 1029, 1033 (D.C. 2013) (quoting Russell v.

United States, 65 A.3d 1172, 1184 (D.C. 2013)). My colleagues conclude otherwise,

ruling that no rational factfinder could draw that conclusion. I disagree and dissent. 28

There was considerable record evidence supporting the trial court’s guilty

verdict for misdemeanor cruelty to animals, beyond the barest facts noted above

(which alone strike me as sufficient to support the guilty verdict). All of the relevant

witnesses and responders, other than Ross herself, recognized that leaving a dog in

a car on a blazing hot summer day—even with the windows cracked a few inches,

and even if only for a far shorter period than it turned out to be—required an

immediate response to avoid harm to the dog. A bystander, Zachary Vasile, was the

first on the scene. He heard the dog’s “very loud” and “incessant barking,”

explaining that—aside from it being “very hot” that day—it was “[t]he constantness

of” the barking that “gave [him] concern.” Vasile understood that “this was a

potentially dangerous situation for the dog.” He spent several minutes trying to track

down the dog’s owner, then tried but failed to open the car’s doors (risking at least

a nasty confrontation on the owner’s return). When that didn’t work, he called 311.

The 311 receptionist then “put [Vasile] to a dispatcher” who put out “a level

1 emergency call”—meaning the situation posed a “risk of death or great bodily

injury”—and dispatched “fire, animal control, and police” to the scene in light of the

temperature that day. At the time of that dispatch Ross was still an hour away from

returning to her car. Firefighters arrived first, with several police and animal control

officers arriving shortly thereafter. The firefighters acted quickly: one tried to

unlock the driver’s side door by attempting to reach through its cracked window, 29

and when his arm wouldn’t fit, he manually forced the window down so that they

could unlock the door and free the dog.

Animal control officer Aristides Torres then arrived on the scene shortly after

Cinnamon had been removed, and he placed the dog in the back of his animal control

vehicle (equipped with dog crates). Video footage of the dog shortly after she had

been freed from the car and placed in Torres’s animal control vehicle (at 5:29 p.m.)

showed her panting heavily, and it would still be another forty-plus minutes before

Ross returned to her car.

Torres provided the most critical testimony in support of the verdict. Torres

testified that in his eight years on the job he had responded to calls for dogs locked

in cars more than fifty times, largely in the summer months. He explained that the

temperature inside of a car can “increase [to] double the temperature” of the external

temperature, “[d]epending on the time” it is parked. And he also explained that it

was “important to get [Cinnamon] out of the car” when the firefighters did because

otherwise “it could definitely have been fatal.” My colleagues posit that Torres’s

testimony in this respect “rests on . . . assumptions,” ante at 18, but eight years as an

animal control officer is called experience, not assumptions. And if Ross wanted to

cast some doubt on Torres’s testimony or test its foundations, cross-examination was

the time to do it (not in a sufficiency appeal); I can understand why defense counsel 30

didn’t want to do that given how readily a longtime animal control officer could

presumably shore up that opinion and strengthen the government’s case. Ross also

had an opportunity to offer any contradictory evidence in the defense case, but failed

to do so. The trial court could thus reasonably rely on Torres’s testimony as ironclad

support for the conclusion that Ross had put her dog in harm’s way.

All of that evidence was further supplemented by a heavy dose of common

sense. Factfinders “need not check their common sense at the courthouse doors,”

and “are permitted to use the saving grace of common sense and their everyday

experience to draw reasonable inferences from the evidence presented.” Covington

v. United States, 278 A.3d 90, 99 (D.C. 2022) (internal marks omitted) (quoting

Long v. United States, 156 A.3d 698, 714 (D.C. 2017)). They are not tabula rasa

bots whose only inputs are the evidence in the trial record; they can bring their prior

knowledge and experiences to bear on their verdicts. And it is common knowledge

that parked cars can quickly get far hotter than the ambient temperatures outside (just

as Torres testified)—most people have opened a car door on a hot day to a blast of

heat and a scorching steering wheel and belt buckles.1 It is similarly common

knowledge that dogs with “dense” fur like Cinnamon are particularly vulnerable to

1 As a born-and-raised Arizonan, I’ve surely had more frequent, vivid, and painful experiences with this phenomenon than most people. 31

heat. While I’m quite confident that any human would suffer if left in a car for an

hour and twenty minutes in ninety-eight degree heat, if you put a dense fur coat on

them, my confidence skyrockets.

Based on all of the above evidence, and the common sense inferences from it,

the trial court quite rationally concluded beyond a reasonable doubt that Ross did

not provide Cinnamon with “proper shelter or protection from the weather,” a

misdemeanor offense per D.C. Code § 22-1001. Really, who could doubt that? As

the trial court opined, “you don’t need expert testimony to conclude that leaving a

dog in a hot car for at least an hour and 20 minutes is failing to provide the dog with

protection . . . from the weather.”

The majority offers four main responses, but they are unpersuasive.

First, the majority suggests that maybe cars heat up to dangerous temperatures

only when left “in full sun” with windows fully up, hypothesizing that the interior of

Ross’s car might actually have been cooler than outside because her windows were

cracked and there was some light shade. Ante at 12-13. I invite them to test out that

theory on the next ninety-something degree day in the District, but there is no need

to dissect the fallacies underlying it here. Suffice it to say that nobody has ever left

their car parked on the street for an extended period on a hot day and then opened

their car door to a gust of cool refreshing air, no matter how much shade they parked 32

in. Some shade or cracked windows might slightly ameliorate how quickly and

drastically a car heats up, but they will not reverse the effect. While the majority

correctly notes that a defendant is not required to prove anything, neither is the

government “required to negate every possible inference of innocence before an

accused may be found guilty of an offense beyond a reasonable doubt.” Hebron v.

United States, 837 A.2d 910, 911-12 (D.C. 2003) (en banc) (quoting Smith v. United

States, 809 A.2d 1216, 1221 (D.C. 2002)). Prosecutors do not need to debunk every

imaginative hypothesis one might later dream up.

The majority also fails to view the record in the light most favorable to the

government, as we are required to do in this posture, when it posits that Ross parked

in the shade. The video evidence shows otherwise. It shows Ross’s car facing

directly into the beating sun, leaving her windshield to act as a giant magnifying

glass for the sun’s rays. There was at most a light smattering of shade hitting some

parts of her car, as depicted in the appendices to this opinion, which are still images

taken from the video exhibits.2 And most importantly, the government produced an

2 Appendix A shows Ross and the conditions around her car when she and her companion returned to it at 6:11 p.m. Notice that the sun is hitting the street directly, with no shade, all around Ross’s car (her car, her companion, and Ross herself are casting the only shadows on the street around her car, though there are some reflections and glare on her car hood that could be mistaken for shadows). Appendix B comes from the vantage point of Torres’s animal control vehicle, which was parked front bumper to front bumper with Ross’s car, and it shows the tree that is 33

experienced animal control officer who was on the scene and thus familiar with all

of those relevant conditions, and he opined without refute that Ross’s dog might

have died had she not been removed from the car some forty-five minutes before

Ross’s eventual return. That is enough to conclude beyond a reasonable doubt that

Ross’s actions created a plain and strong likelihood that Cinnamon would be harmed

but for responders’ interventions.

Second, the majority invokes a countervailing principle to the one I have

relied upon above, that factfinders can bring their common sense to bear on verdicts.

The majority counters—citing to Mejia Cortez v. United States, 256 A.3d 210 (D.C.

2021), and Long, 156 A.3d at 714—that “even where the fact in question may seem

incontrovertible,” “inferences and common sense cannot serve as substitutes for

evidence.” Ante at 10. That generic principle is too broad if taken at face value;

common sense can obviously fill some evidentiary gaps in a record, as our

the only candidate for providing some shade to Ross’s car. The reader will notice how its shadow is being cast away from the street and onto the nearby building. Some stray branches may have jutted out enough to lend some light shade to Ross’s car, but it wouldn’t have been much. Appendix C confirms the point. It shows the sun very directly hitting the back of Torres’s animal control vehicle (i.e., roughly as it would have hit Ross’s front windshield). The reader will also notice the direct and bright glare of the sun coming straight through the light tree cover. Further notice in Appendix C that while the trees and houses on the opposite side of the street are casting shadows into the street, none of the visible trees on Ross’s side of the street are doing so because they are of course casting shadows in the same direction— away from Ross’s side of the street. Even when viewed generously to Ross, Ross’s car was receiving only a light smattering of shade from a nearby tree. 34

precedents make clear. For instance, there was no direct evidence at trial that a dog

is an “animal” covered by the animal cruelty statute, but I assume the majority agrees

with me that common sense can supply that link. And if Ross had intentionally

broken each of Cinnamon’s legs with a baseball bat, or stabbed her in the abdomen

with a buck knife, I hope the majority agrees that no further proof would be

necessary to establish a plain and strong likelihood that those actions would harm

the dog (despite its rhetoric that “[r]es ipsa loquitur” has no role to play in the

criminal law).

The question always comes down to whether the particular fact at issue is truly

common sense enough to be inferred from the record evidence, and both Mejia-

Cortez and Long quite sensibly answered in the negative on their particular facts.

The disputed fact at issue in Mejia-Cortez was whether the Washington Metro and

Transit Authority “had been issued a license to sell alcoholic beverages,” and in that

case the government did not even “suggest that it could satisfy its burden by pointing

to the obviousness” of that fact. 256 A.3d at 214-15. In Long the particular fact was

whether a decade-old 2002 Dodge Intrepid was worth more than $1000 when there

was no direct evidence of its value and no circumstantial evidence about its mileage,

maintenance, or mechanical condition—aside from it being operable—factors which

could obviously drastically affect its value. 156 A.3d at 714. While we held that

one could not rationally conclude beyond a reasonable doubt that the car was worth 35

more than $1000 on that record, as necessary to support a felony receipt of stolen

property offense, we directed entry of a misdemeanor receipt of stolen property

conviction, because of course the car had some value despite no direct evidence of

that. We also acknowledged our prior holding that one could rationally conclude

beyond a reasonable doubt that a relatively new car “clearly had a fair market value

of more than $250,” even absent direct evidence on the point. Id. at 714-15

(discussing Curtis v. United States, 611 A.2d 51 (D.C. 1992)).

Both Mejia-Cortez and Long strike me as sensible because the disputed facts

in those cases were not nearly so evident as the challenged fact here—that leaving a

dog in a car for an hour and twenty minutes in ninety-eight degree heat creates a

plain and strong likelihood of some harm to the animal. Had the government charged

Ross with felony cruelty to animals, which would have required proof of a

“substantial risk of death” or serious bodily injury, D.C. Code § 22-1001(c)-(d), I

might agree that the evidence was insufficient to prove that aggravated five-year

felony offense. But Torres’s testimony that Cinnamon might have died had she not

been rescued from that car forty-five minutes before Ross’s return was itself a

sufficient basis on which to conclude that Ross’s actions created a plain and strong

likelihood of some harm to the animal, even if only some unquantifiable degree of

suffering. Jordan v. United States, 269 A.2d 848, 849 (D.C. 1970) (examining

whether dog was left in conditions that would “cause [it] to suffer” under animal 36

cruelty statute). That is all that the government was required to prove for the

misdemeanor offense that Ross was convicted of. Id.

Third, the majority relies on Jordan, id., in which we reversed an animal

cruelty conviction due to a lack of sufficient evidence to support it. Ante at 8, 19

n.8. But again, the particular facts matter a great deal to any sufficiency analysis,

and the majority omits them. The facts in Jordan were that a full-grown German

shepherd was left outside in roughly twenty-five degree temperatures for about five

hours. We stressed that “[i]t is a matter of common knowledge that some breeds of

dogs can remain exposed to extremely cold weather for many hours without

injurious effects,”3 269 A.2d at 849, as anybody who’s even roughly familiar with

the Iditarod knows. There is no corollary common knowledge—because it is

counterfactual—that there are dog breeds that can be safely left in a car for an hour

and twenty minutes in ninety-eight degree heat. They do not run a Death Valley

summer Iditarod. So Jordan is of no help to the majority.

To be sure, a person can always be wrong in what they think to be common

knowledge, so that a factfinder should always be willing to reexamine and abandon

their priors in light of the evidence presented at trial. And an appeal to common

3 It strikes me as reasonably well known that German shepherds are one of those dog breeds that endure the cold well, or at least that is fairly inferable from their heavy coats. 37

knowledge generally cannot stand up against evidence that actually undermines or

contradicts it. But as I have detailed above, the evidence at trial only corroborated

that Cinnamon was placed in a harmful and potentially deadly situation, so there was

nothing irrational about the trial court’s verdict.

Let me nonetheless briefly scrutinize what I have claimed to be common

knowledge, using both the relevant facts of this case and those presented in Jordan.

I have asked ChatGPT, “Is it harmful to leave a dog in a car, with the windows down

a few inches, for an hour and twenty minutes when it’s 98 degrees outside?” 4 Its

response, with my emphasis added, follows:

Yes, leaving a dog in a car under these conditions is very harmful. Even with the windows slightly open, the temperature inside a car can quickly rise to dangerous levels. In fact, on a 98-degree day, the

4 I do not mean to suggest that ChatGPT is a good proxy for what is, and isn’t, common knowledge—it is definitely not. It can answer questions that are far from common knowledge. For instance, when I asked it what the forty-first element in the periodic table is, it responded (apparently correctly) with Niobium (Nb). But it is at least a point in favor of something being common knowledge that it is, in fact, true; and it is a far stronger point against it if it’s not. Also, when I had each of my three law clerks run this same inquiry, ChatGPT gave each of them a slightly different but substantively similar answer, all beginning with an unequivocal “Yes.” Because I am writing only for myself, I will not delve into the many issues and concerns surrounding using AI-powered large language models like ChatGPT in judicial decisionmaking. If this dissenting opinion had some precedential force, I would at least want to explore the value and dangers of using AI even when used only for this bit of color commentary. But for present purposes, I am content to point the reader to a thoughtful and engaging discussion on that topic, which I largely agree with. See Snell v. United Specialty Ins. Co., 102 F.4th 1208, 1221-35 (11th Cir. 2024) (Newsom, J., concurring). 38

temperature inside a car can escalate to over 120°F (49°C) in just a matter of minutes, which can cause heatstroke or even be fatal to a dog.

Dogs can suffer from heatstroke when they are exposed to extreme temperatures, as they do not regulate heat as efficiently as humans. They rely on panting to cool down, but this becomes less effective in a hot, confined space like a car. Symptoms of heatstroke in dogs include excessive panting, drooling, weakness, vomiting, and even collapse.

For your dog’s safety and well-being, it’s important never to leave them in a hot car, even with the windows cracked. If you need to run errands or be in a hot environment, it’s best to leave your dog at home in a cooler, safer space.

Now compare that unequivocal affirmative answer to how it responds to the

facts of Jordan. I asked ChatGPT, “Is it harmful to leave a German shepherd

outside in 25 degree temperature for five hours?” The first paragraph of its far

lengthier response—which boils down to “it depends”—was:

Whether it’s harmful to leave a German Shepherd outside in 25°F (- 4°C) for five hours depends on several factors, including the dog’s health, coat condition, and access to proper shelter. German Shepherds are a hardy breed with a double coat that provides some insulation, but prolonged exposure to cold can still be harmful. Here’s what to consider:

It then details five relevant factors that should be taken into consideration, including

the dog’s “health and age,” its “coat condition,” its “activity level,” its access to

shelter, and the “duration” (despite five hours being baked into the prompt). 5

5 The majority suggests that ChatGPT’s answer to its own question about the value of a 2002 Dodge Intrepid in 2012 is in some tension with Long, but it isn’t. 39

If I were to reframe ChatGPT’s answers in terms of the relevant legal

standards, its first answer reads to me as something like, “Yes, beyond a reasonable

doubt, leaving a dog in a car for an hour and twenty minutes in 98-degree

temperature is very likely to cause it harm.” Its second answer, concerning Jordan,

is “you could not say beyond a reasonable doubt that leaving a German shepherd

outside in 25 degree temperature is likely to cause it harm, though it possibly could

depending on a host of factors.” I think that aligns perfectly with what my own

common sense tells me—and at least a factfinder would not be irrational in tracking

those lines of thought—so that we rightly reversed the conviction in Jordan, while

Ross’s conviction should be affirmed. But I digress.

Fourth, the majority seeks to distinguish a few analogous cases from other

jurisdictions where animal cruelty convictions were upheld despite the dogs in each

Using the majority’s own Q&A, ChatGPT answered that the Intrepid’s value would “likely” fall in the $3000 to $5000 range, but it noted that the very same factors we highlighted in Long—mileage and maintenance—could bring it outside of that range. By simply asking a more targeted question of ChatGPT, it confirms the point: “Would you say, beyond a reasonable doubt, that an operable 2002 Dodge Intrepid would be worth more than $1000 in 2012?” Its answer is roughly the same one we gave in Long. To paraphrase, it says that the car would “likely” be worth more than $1000, “but whether it was beyond a reasonable doubt depends on factors like condition, mileage, location, and market trends at the time.” Conversely, when I ask whether it can “say, beyond a reasonable doubt, that leaving a dog in a car for an hour and twenty minutes in 98 degree heat would raise a plain and strong likelihood of harming the dog,” its answer is “Yes, beyond a reasonable doubt,” with extensive elaboration. 40

case being left in cars for both briefer periods and on cooler days. See City of

Beachwood v. Pearl 111 N.E.3d 620, 636-37 (Ohio Ct. App. 2018) (forty minutes

in 84 degree temperature); State v. Washburn, 325 A.3d 136, 139 (Vt. 2024) (more

than “five to ten minutes” after an officer saw the dog in “moderate distress” in 78

degree temperature); State v. Butler, 293 A.3d 191, 193-94 (N.H. 2022) (thirty

minutes to an hour with the temperature “around 92 degrees outside”). Notably, it

offers no analogous cases of its own where animal cruelty convictions were reversed

on even roughly similar facts.

The majority attempts to distinguish the above cases on the basis that there

was some evidence that the animal was in actual distress in each of them, whereas

here, Torres opined that Cinnamon “looked fine” and did not appear “in distress”

when he arrived on the scene. What the majority overlooks, or at least drastically

downplays, is that Ross was not even halfway into her frolic when Cinnamon was

rescued and Torres arrived on the scene—Cinnamon was freed roughly thirty-five

minutes into Ross’s hour-and-twenty minute absence. So the fact that Cinnamon did

not appear to be in distress yet, shortly after being removed from Ross’s car, hardly

undercuts the conclusion that there was a strong likelihood that she would get there

over the next forty-five minutes had she not been rescued (which is still longer than

two of the cases cited above, and on par with the third). In fact, Torres testified that

“excessive[] panting” is an early sign of distress in a dog, and video evidence taken 41

several minutes after Cinnamon was removed from the car showed her to still be

panting heavily. While Torres did not seem to think Cinnamon’s heavy panting

alone amounted to actual distress, it is certainly some support for the conclusion that

she was well on her way, less than halfway into Ross’s absence.

So whatever one thinks about what I have claimed to be common knowledge

above, the actual and uncontradicted evidence in this case—taken in the light most

favorable to the verdict—was that the temperature inside of Ross’s car would have

been much hotter than the 98 degrees outside.6 And if firefighters did not rescue

Cinnamon from the car when they did, “it could definitely have been fatal.” That is

more than a sufficient evidentiary basis for a factfinder to rationally conclude that

Ross’s actions created a “plain and strong likelihood” that Cinnamon would be

harmed, and so I dissent. I would affirm Ross’s conviction.

6 While Torres said more specifically that the temperature could double inside a car depending on the time it was parked, he did not specify how much time it would take for that to occur. So I do not think one could fairly infer from his testimony that the temperatures in Ross’s car might have reached double the outside temperatures (196°F) in the time she was gone. Could the temperatures in that car have gotten about halfway there, and reached into the 140°s? Sure, I would say that’s within the realm of a reasonable inference that one could rationally draw in the government’s favor on this record. But it is safer still to say only that the temperatures would have been much hotter in that car than outside of it, and I think that is sufficient to support the conviction in this case.