Tesla has released its Autopilot-related safety report for the first quarter of 2021 and the results have prompted Elon Musk to claim a Tesla with Autopilot has a 10 times lower chance of an accident than the average vehicle. That, however, is not the full story.

In its safety report, Tesla revealed that it registered one accident for every 4.19 million miles driven by owners with Autopilot engaged. For those driving without Autopilot but other active safety features, there was one accident registered for every 2.05 million miles driven. Those driving a Tesla with neither Autopilot nor active safety features registered one accident for every 978,000 miles driven.

Read Also: No One Was Driving The Tesla Model S In Fiery Crash Where Two People Were Killed

The car manufacturer noted that there is an automobile crash every 484,000 miles driven in the U.S., according to the National Highway Traffic Safety Administration. This prompted Musk, and many Tesla enthusiast websites, to essentially proclaim that a Tesla running on Autopilot is 10 times less accident-prone than the average vehicle.

What Tesla fails to mention is the fact that Autopilot is predominantly used by owners for highway driving but NHTSA crash data doesn’t break down crash rates on highways. As Electrek points out, accidents are more common on city roads and undivided roads, meaning Tesla’s comparison to NHTSA crash data is not really valid.

In addition, this latest crash data released by Tesla is actually worse than in Q1 2020. In fact, in the first quarter of last year, Tesla recorded one accident for every 4.68 million miles driven with Autopilot.

Where improvements were made were with Autopilot disengaged but active safety features enabled as one accident was recorded every 2.05 million miles in Q1 2021 compared to every 1.99 million miles in Q1 2020. In addition, Tesla’s average of one accident for every 978,000 miles driven in Q1 2021 without Autopilot or active safety features is down on the average of one accident for every 1.42 million miles driven in Q1 2020.

So, what does this all mean? It’s simple, really: as Electrek puts it, the data is so incomplete that they’re not useful enough to make comparisons.