77% of New Cars Can Detect Motorcycles. The Other 270 Million Vehicles Can’t.

IIHS dropped a motorcycle-shaped foam target in front of new cars this year, ran them at it at 43 mph, and graded the results. Most stopped. The Institute announced in March 2026 that its updated crash prevention evaluation now includes motorcycle and semitrailer targets at three speeds (31, 37, and 43 mph) and that passing is required for the Top Safety Pick+ award.^[1] More than 77% of 2026 model year vehicles tested earned acceptable or good ratings.

Good for the 2026 models. Bad for the roughly 283 million registered vehicles already on American roads that can’t detect a motorcycle if you parked one on their hood.^[3]

The body count nobody stopped

FARS recorded 6,335 motorcycle fatalities in 2023. Not a typo, not a rounding error. The actual highest single-year motorcycle death toll in U.S. history.^[2] A 26% increase since 2019. Motorcyclists now account for 15% of all traffic deaths, up from 7% in 1975. Their share has more than doubled while car occupant deaths plateaued and dropped.

Sixty-five percent of those deaths — 4,118 riders — occurred in crashes involving at least one other vehicle.^[2] A car, a truck, an SUV. Something with a bumper and a driver who didn’t see them.

AEB technology that detects motorcycles exists right now. IIHS just proved most new cars have it. Question is what happens during the two decades it takes for those new cars to replace everything else.

Fleet math

Average age of a light vehicle on U.S. roads hit 12.6 years in 2024, per S&P Global Mobility.^[3] That means the median car at any given intersection is a 2012 model. AEB of any kind was a niche luxury option in 2012. Motorcycle-specific AEB started appearing around 2022.

Call it 5–7% of the current fleet that’s new enough to plausibly have motorcycle-capable AEB. Multiply by the 77% pass rate. Roughly 4–5% of cars on the road today could theoretically detect and brake for a motorcycle.

NHTSA’s AEB mandate kicks in September 2029.^[4] After that, every new light vehicle sold must have it. But “every new vehicle” is not “every vehicle.” At current fleet replacement rates, 15–17 years pass from a mandate before the technology reaches even 90% of on-road cars. ESC was mandated in 2012 and still hasn’t fully saturated.^[5]

Motorcycle-detecting AEB won’t reach the majority of cars until roughly 2040. Near-universal by 2045.

How many riders die waiting

Back-of-envelope, assumptions laid bare:

• 4,118 annual multi-vehicle motorcycle deaths (2023 figure)
• 30% estimated crash avoidance rate for motorcycle-specific AEB, extrapolated from IIHS rear-end AEB studies showing ~50% reduction, discounted because motorcycle crashes involve more turning conflicts and higher closure speeds^[6]
• ~5% current fleet penetration of motorcycle-capable AEB

Today: 4,118 × 5% × 30% = ~62 lives saved per year.

Full fleet: 4,118 × 30% = ~1,235 lives saved per year.

We’re capturing roughly 5% of what the technology could deliver. Ninety-five percent of the theoretically preventable deaths are happening on schedule because the fleet is old.

Across the 19 years between now and approximate full saturation, at current motorcycle death growth rates, something on the order of 75,000–80,000 motorcycle riders will die in multi-vehicle crashes. Many in the exact scenario 77% of 2026 models can already prevent.

Strongest counterargument

Motorcycle rider behavior is the elephant on two wheels. Thirty-five percent of fatally injured motorcycle drivers in 2023 had no valid license.^[2] Alcohol involvement is persistent. Single-vehicle crashes (35% of motorcycle deaths) are by definition immune to car AEB. Even in multi-vehicle crashes, riders are frequently speeding, lane-splitting, or running lights. Investment in licensing enforcement, helmet laws (only 17 states require universal helmet use), and basic training might deliver faster returns per dollar than waiting two decades for fleet turnover.

Partially correct. Rider behavior is a massive factor. But it doesn’t help the 4,118 killed in multi-vehicle crashes, many of whom were doing everything right when a driver making a left turn looked straight through them. AEB doesn’t care whether the rider has a license. It brakes for the silhouette.

Limitations

Our 30% effectiveness estimate is this article’s weakest link. IIHS AEB studies focus on rear-end car-to-car crashes; motorcycle crash dynamics are fundamentally different. No peer-reviewed study has measured motorcycle-specific AEB effectiveness at scale, because the technology is too new. Fleet penetration (~5%) is derived from new vehicle sales fractions and average fleet age, not VIN-level AEB installation data. Multi-vehicle death figures include cases where the motorcycle struck the car, where AEB on the car would have been irrelevant. And motorcycle fatality trends may not continue upward indefinitely; gas prices, EV motorcycle adoption, and future licensing reforms could bend the curve.

FARS captures only fatal crashes. Non-fatal motorcycle injuries, hundreds of thousands annually, many catastrophic, are excluded from this analysis.

Where this leaves us

IIHS proved the engineering problem of detecting motorcycles with camera-and-radar is solved. Three out of four new cars can do it. Between now and the day most cars on the road actually have it, approximately 80,000 riders will die in crashes with vehicles that couldn’t see them.

IIHS President David Harkey: “The safest crash is the one that never happens.”^[1] Also true: the deadliest safety technology is the one sitting in a showroom while 2012 Camrys run left turns through motorcyclists on unlit suburban arterials.