Did you mean commercially, or just any experimental prototype flying any distance?

@WilliamGunn it’s not required to be done commercially, but it does need to be something documented and verifiable. If the experiment or prototype has a way for us to confirm it, especially video footage or coverage by journalists, it absolutely would meet the criteria.

Airbus did an autonomous ATTOL program in 2020. Although it’s unclear if there were any flights with no pilot input and a pilot aboard.

Every spacecraft ever flown has been autonomous. It’s pretty amazing that this isn’t already the case for aircraft.

Isn't there some VTOL company that's planning on doing this?

If there is a human pilot in charge of overseeing the AI potentially taking over in case there is an issue, but otherwise the AI is in charge, how would you resolve?

@Odoacre would it make a difference if the pilot themselves are on board or not?

What counts as "AI" here? I feel like in principle you can probably do this already without using any fancy ML. And are there any requirements on the passengers? Like is it enough to have a person on the aircraft, or does it have to be a commercial flight, etc.

@jskf I tend to agree that in theory current technology may be enough to accomplish the task today. To my knowledge the more difficult portion is in the doing of the task.

Obviously autopilot and similar things can perform most functions, but I have been unable to find any evidence that an entire flight including takeoff and landing has been performed with human life at risk.

The question is whether or not it will be attempted. I hope that clarifies the market.

@Dustin If there has been a 100% autopilot flight already, does that resolve yes or n/a?

How many human passengers? At least two?

What type of aircraft? Plane, helicopter, drone, blimp, ...?