You always have the password as backup to unlock. If your appearance were to suddenly change remarkably, such as via plastic surgery or a horrific accident, you would set your new face after unlocking with the password.

Based on their phrasing, I'm assuming the "adaptive" aspect means that slow changes, such as gaining or losing weight, will progressively train the existing face - without it ever failing to recognize you because you passed some weight threshold. Same would go for growing a mustache or beard. As to whether you can shave off a 5-year beard and have it still recognize you without manual intervention, that would depend on how much the algorithm cares about your chin and lip regions vs. the rest of your face. Based on the demo, the hair on top of your head is 100% ignored.

Apple claims that the odds of a random person's finger unlocking your iPhone with Touch ID is 1 in 50,000, and the odds of a random person's face unlocking Face ID is 1 in 1,000,000.

That's interesting. I would think that identical twins would be very hard to distinguish... and those certainly have an almost 1:350 odds. Of course you're likely to know them, but still I'm not sure banks would be happy with that for FIDO.

They did say to not enable Face ID if you have an "evil" twin, which is pretty jokey, but probably 100% true.

A twin could walk into any bank with their sibling's passport and do anything they wanted.

Banks are all about risk management not necessarily avoiding and and all risks.

But a twin wouldn't be able to unlock my iPhone 7 with their finerprint.

You just disable FaceID in that scenario and go back to using a passcode.

But again if you can't trust your twin well then your phone is the last thing I would be worried about.

I think the number is if you choose another human at random, how likely is it that they'll be able to unlock your phone. Having a twin doesn't change that number much, since they're only one person out of 7+ billion. Of course, your threat model may be different from "pick a random human from anywhere on the planet."

> I would think that identical twins would be very hard to distinguish

Ask their mom. (Hint: They're not.)

If you want to teach Face ID to reject masks, you need to make some masks. Similarly, if it needs to be taught to reject a twin, you need dozens of twins. And if it starts labelling people incorrectly as their twins, is it worth it?

Perhaps they can sidestep this by offering a specific twin learning feature.

Twins are just examples of two people with very similar faces. If Apple are able to train Face ID to distinguish between 6 billion different faces they will also be able to distinguish faces of twins.

The twin's mother is inside the phone telling it what to do? Or the person you replied to is talking about technology finding it difficult to distinguish, and not a close family member?

My point is that twins are only difficult to distinguish for people who do not know them. For family and friends it's easy, which means that there are substantial differences even for faces of twins. The face detection technology will be able to recognize those differences as good or better than the twins' family and friends. If it does not now, it will eventually.

I believe it uses an IR camera, or similar tech, defeat that, sensing depth. Kinect style. A photo won't defeat it.

This time they’ve got an infrared sensore, as well as two cameras with different focal lengths, so they should be fine.