Can AI Catch a Lame Horse Before You See It?

Last updated: September 28, 2026

By: Miles HenryFact Checked

If you’ve ever had a horse that’s “just not quite right,” you know how frustrating it is when everyone sees something different, and nobody can put a finger on the problem. AI lameness detection in horses gives owners and vets another way to measure small movement changes that can be hard to see with the naked eye.

Here’s what horse owners should know: how these tools work, the main systems vets and racetracks use, what the research actually shows, what AI can’t do, and how to use the data without getting ahead of your vet.

How AI Lameness Detection Works

If you’ve owned enough horses, you know the hardest cases aren’t the obvious ones. An obviously three-legged-lame horse is usually easy to spot. It’s the one that’s a little short-strided one day and looks fine the next that keeps owners and vets chasing answers.

A sound horse generally moves evenly from stride to stride. When a horse protects a sore area, even a small change in how he loads a leg shows up in the way his head, withers, or pelvis move. During a lameness exam, a vet watches for changes like a head nod or uneven hip movement when the horse jogs. AI tools measure them.

Right now, most systems fall into two categories:

• Video-based tools use a smartphone recording of the horse trotting. The software tracks points on the horse’s body frame by frame and calculates how uneven the movement is, without any markers or sensors on the horse.
• Sensor-based tools use small motion sensors, called inertial measurement units, attached to the horse’s head, body, legs, or saddle cloth. They record movement many times per second and flag patterns that differ from a sound stride.

The Main Tools

Features and pricing change often, so check each company’s site for current details. Here’s how the best-known systems differ:

Racing has become an important testing ground for this technology. According to StrideSAFE, its sensor is deployed at every racetrack in Kentucky, where trainers can see their horses’ reports through a confidential portal. And the British Horseracing Authority has trialed Sleip during pre-race vet exams at British race meetings, including Aintree.

What the Research Shows

• Smartphone video can measure asymmetry at clinically useful levels. In a peer-reviewed validation study of 25 horses, researchers compared Sleip’s smartphone system with multi-camera motion capture and found a mean difference of about 2.2 millimeters for head and pelvis movement. In plain English, a properly recorded phone video can give your vet objective movement measurements to weigh alongside the exam.
• Sensors can flag racehorses at higher risk. A Washington State University-led study in the Journal of the American Veterinary Medical Association tracked StrideSAFE data from nearly 30,000 races and more than 11,800 Thoroughbreds at 10 U.S. tracks. Horses given the highest risk score were 44 times more likely to suffer a fatal injury than those with the lowest. Some of the authors are StrideSAFE shareholders. The practical takeaway is that fine-grained stride data may point to horses that deserve a closer look before a serious problem becomes obvious.
• The signal appeared across multiple sensor companies. In the AAEP’s prospective study of 561 2-year-old Thoroughbreds during the 2025 season, horses flagged yellow or red after a breeze were about twice as likely to have a musculoskeletal injury as horses flagged green. That’s encouraging, but it still means “take a closer look,” not “this horse will get hurt.”

These tools raise or lower the odds. They don’t tell you for certain that a horse will get hurt, and they don’t tell you what’s wrong.

What AI Can’t Do

• It can’t find the cause. An app can tell you the right front looks off. It can’t tell you whether that’s a foot bruise, a sore fetlock, or a suspensory problem. Nerve blocks, imaging, and a vet’s hands still do that. My guide to horse lameness causes and treatment covers the usual suspects.
• Asymmetry isn’t automatically lameness. Horses can show small, repeatable movement differences without an obvious clinical problem. That’s why a baseline, simply knowing what normal looks like for your horse, often tells you more than one isolated reading.
• Bad data gives bad answers. Uneven ground, a horse that won’t trot straight, a camera held at the wrong angle, or a sensor that shifts can all throw off the results.
• It’s a screen, not a diagnosis. A flag tells the people responsible for the horse that something may deserve a closer look. It doesn’t, by itself, mean a horse is injured or should be scratched. As Dr. Larry Bramlage put it when discussing the AAEP sensor study, the sensors aren’t meant to scratch horses.

How Owners Can Use It

Most owners don’t need a daily gait score on a horse that’s moving normally. These tools are most useful when you have a baseline from a sound day, a horse that’s hard to call, or a rehab case where you and your vet want to see whether the trend is really improving.

1. Get a baseline while the horse is sound, so later readings have something to compare against.
2. Record the same way every time: same level, firm surface, same handler, and a straight line at a steady trot. Follow the app’s instructions for camera distance and angle.
3. Share the results with your vet rather than acting on them alone.
4. Use it to track rehab, so you can see whether a horse is actually improving before you and your vet decide to increase his workload.

When Movement Data Actually Helps

Say your horse isn’t obviously lame, but your trainer says he feels a little different behind. You don’t want to turn him out for weeks based on a hunch, but you also don’t want to keep training through a developing problem. A consistent movement record gives your vet another objective comparison point alongside the exam, the horse’s history, and any changes in his workload. It may support a decision to monitor closely, modify his work, or investigate further.

If you want a checklist to take to the barn, my printable AI lameness detection checklist covers the questions to ask before you pick a tool.

AI Lameness Detection in Horses FAQs

AI is getting better at spotting small movement changes before they become obvious to the naked eye. That doesn’t replace good horsemanship, a consistent routine, or a thorough vet exam. But when you know what normal looks like for your horse, movement data can help you notice a change sooner and start the right conversation with your trainer and vet.