The internet is overloaded with AI slop and terrible videos of “paranormal” fakery. Once in a while, an interesting one comes along. I found this today on TikTok. It’s posted by a normal person who has horses. She has a few security cameras around. On September 2, 2026 she got this video of what has been interpreted as a “ghost girl” visiting a horse at 11pm on Monday night.
She noted additional details in a followup video including that the camera only records in 20 second clips when it detects motion. But was it triggered by the horse moving or a person around it? The other motion sensitive cameras in this area didn’t show anything. I have a similar trail/game type of camera and it does miss some images if the motion is too quick or too far away. The property is fairly isolated but it’s possible that a real child from a neighborhood came around. If so, that would also be weird. This looks to be a girl, 4 or 5 years old, in a nightgown or dress.
The followup video provides the scene in daylight so you can see the other objects in the video.
The best guess about what is going on here is that it is a real person, which is, itself, disturbing. But there are no other leads regarding who it could be.
Some people will say AI, but that doesn’t feel right for this person. While some hoaxes or normal explanations are SO obvious that they hit you over the head, others are not. It isn’t helpful to dismiss it as “hoax” without some justification. I have been working on a post about other “terrible skeptical explanations” that I will finish up in a bit. The point is, if you wish to give your opinion, it should be an educated one. In this case, and in others I’ve posted, I either have some clue about what happened (which is often wrong), or I just don’t know how or why something happened.
I don’t actually find typical ghost claims and research interesting anymore because the evidence is so poor or can’t be reproduced. So what good is it? At this point, it’s about belief. Unless we have some unexpected tech breakthough to explain space-time anomalies, we just aren’t getting anywhere with the big questions about paranormal claims.
Kenny Biddle should see this. Is he on your mailing list? He’s adept at analyzing photo and video evidence.
The original video has many blurry streaks moving from upper right to lower left, as if out of focus leaves, small insects, or raindrops are being blown across the foreground by the wind. The horse’s tail, however, hangs straight down. To me, this suggests that two videos have been combined. If you were in possession of the first definitive evidence for life after death, would you announce it to the world by splicing in some cheesy spook-music and posting it on Instagram?
It was TikTok in this case. But maybe. What else would they do with it. It’s a way to get it out to people. She isn’t saying she knows what it is. I also noted the insects but I don’t put much weight on there being wind that would suggest a modified video.
The most likely explanation is that there really was a little girl, and then the only remaining question is one of motivation. It wouldn’t be the first time a pre-teen girl snuck into a stable to bring Dobbin a sugar cube. Or she could have been recruited to do her perp-walk, either by the poster herself in search of likes, or by somebody playing a practical joke on the poster.
Please refrain from making assumptions and allegations.
This appears to be a VERY little girl, not a “pre-teen”. To suggest the person “recruited” her for “perp-walk” is a bit offensive.
Again, skeptics making unsupported speculation is unhelpful.
I used the word “likely” advisedly. Those were suggested possible explanations, not “assumptions”, and not “allegations”. Also pre-teen means less than 13 years old, so I assume we can agree on at least this. One other thing I thing I think we can agree on: this is not a ghost, because ghosts don’t exist. Given that, this video was created to give the impression of a ghostly figure visiting a stable in the middle of the night. If that’s an allegation, then so be it. I allege that somebody created this video with an intent to mislead.
Went to a lecture once by a professor of photography. He went through a lot of these ghost/alien photographs and pointed out how they were all artefacts of the camera and the photographic process. This looks a bit different but I’d like to see an analysis by someone who knows photography.
Soooooooooooo what does the rest of the video show? That horse seems awfully unresponsive to the ghost.
As noted, there was only a 20 sec record on the camera. That’s standard for trail cams.
I blanked right over the time thing but I still think the horse is awfully unreactive.
It’s hard to tell what’s happening in this video, and I do think it might show a real person, but there are clearly effects of the high-gain, time-integrating “night vision” sensor at play. These sensors, widely used in surveillance cameras and game cameras, integrate over multiple frames to “amplify” light and produce a more usable image. That means that each frame you see is something like a statistical sum of the frames before it, which can stretch into a period of over a second in really dark conditions.
One effect of these sensors is that dark moving objects often appear transparent, because they are only present in each region of the sensor for part of the integration window and thus get “averaged” together with the background behind them. We can see this happening with a post behind the “girl” at the beginning of the video, which fades away as the girl has been in position in front of it for long enough for the background to fall out of the integration period.
This same trait of these sensors also tends to distort the shape of moving objects in dark conditions. There are a number of well known surveillance camera “alien” videos that depict dogs that appear very strangely (even as bipedal) because they are moving horizontally and the integration window effect “compresses” them on the axis of motion, so that the outlines you see are only the parts of the image that were covered by the dog for hundreds of milliseconds as it moved… smaller and shorter than the original dog. This effect also tends to make limbs look weird and move strangely. There’s definitely some of this happening in the early part of the video where the motion is really weird and hard to interpret.
One theory I would present is that this video shows another animal, maybe a game animal, that looks like a person because of shape distortion. I’ve seen examples of this before. But I don’t think it’s what’s happening here, because in the later part of the video the scene is very static (the girl is hardly moving at all). This means that everything was in position for the whole integration period and the shapes/outlines we end up with should be pretty accurate. This is also the part of the video where the ghost most resembles a girl in a skirt, so I tend towards that actually being what we’re seeing. I am not very confident though and could see this going either way.
One way to intuit about it is like this: night-vision camera sensors are doing a lot of behind-the-scenes work to interpolate or infer what’s in the scene, given very little actual data (light received). Because they present this as a video, we as humans kind of lose track of the amount of inference that has happened and interpret things too literally. Fine details in night-time footage like this are simply not ground truth, they’re a result of both the sensor design and usually some software post-processing trying to interpolate the details, and especially with anything moving they are frequently wrong. You have to treat this kind of video like eyewitness testimony or a sketch, probably generally correct but unreliable in any of the details.
Yeah, I definitely see the distortion and compression artifacts on the objects. That’s what makes this interpretation a problem. I can’t agree that this shows anything like an animal. If anything, that’s a person. It’s clearly walking on two legs, is light colored (unusual for wild animals) and apparently wearing clothes. So, unless you can make a reasonable hypothesis, I don’t think replacing an unknown or unusual with a more unknown and unusual is feasible.
I think people mistake cynicism for skepticism a lot. We don’t need to have a definite answer to be skeptical that it’s a ghost. I like spooky stuff, I like hearing people’s spooky stories. I agree that we should give people the benefit of the doubt before we jump to hoax. I don’t believe in the paranormal. I think this is most likely a little girl, which as you say is worrying it itself.
All I know is: if I ever become a ghost, I am 100% petting all of the animals. I’ll be sneaking in through doggy doors to curl up with your puppies, I’ll be conjuring apples for your horses, I’ll be stealing your hot dogs for raccoons, I’ll be in the zoo with the jungle cats…
Totally! Me too.
I finally got a chance to watch the videos and it really does look like a little girl. It’s hard to tell from the video, but it looks like the girl puts her hand up to the horse, who then turns his head away as though not interested in contact (he’s probably sleepy). I hope it’s not an actual child wandering around by herself at night. I like the practicality of the owner in the follow-up video–and that horse is so well cared for to be 37yo! I made the mistake of looking at comments in the second video and now I’m enraged at the idiots commenting that a ghost was visiting to prepare the horse for his death. I hope that horse lives to be 50 just to show them.