That press release is not evidence that Tesla's Autopilot is saving lives, for at least three major reasons. For one thing, Autopilot is more likely to be enga

That press release is not evidence that Tesla's Autopilot is saving lives, for at least three major reasons. For one thing, Autopilot is more likely to be engaged for some types of driving than others. Comparing statistics with Autopilot engaged to the total of all miles driven is comparing apples and oranges. Similarly, it's comparing statistics about the people who buy and drive Teslas with the entire driving public. You don't know how many accidents Tesla owners would have gotten into if they were in cars that didn't have self-driving features. Finally, you can't draw any conclusions about whether lives were saved by just looking at the number of accidents and ignoring their severity. Hypothetically, if Autopilot caused fewer accidents at low speeds but more catastrophic collisions on highways, it could be easily be causing more deaths than it prevents.