So we all have cameras on our phone, and we all have faces on our bodies(hopefully). We all also have a want to keep our data safe and secure. To that end many companies have tried to make secure facial recognition software or bio metric scans that can't be beaten with a simple picture.
The idea behind it is always the same. We'll use the unique identifying properties of a person to identify them. The idea behind it is pretty solid. However there's always been one problem with it on some of the earlier implementations that were brought out that showed a copy just has to be good enough. With image(facial/retinal) recognition it was beaten with a picture. With full fingerprint scanners it was beaten with silly putty. Now we have new technologies to try and verify them much more "securely".
Microsoft uses Hello, most phones now have a finger scan, and many of these systems with small changes are considered secure. Are they really though? Theoretically most bio metrics can be imitated well enough to fool the scanners with enough time and money.
So this brings the real question into it: How can you confirm they are really that bio metric in front of your scanner?
I've often thought about this and the one theory I've struck upon is using video capture to prove it's a living thing in 3D space(requires two cameras for depth perception), but even then if they wear a good enough mask it can still be fooled.
So how can we ABSOLUTELY make sure that the bio metric we're scanning is the original bio metric and not a good enough recreation out of clay?
Edit: I'm not asking why it's not used much, I'm asking for theoretical ways to make them as secure as possible through changes to implementation. How we currently take steps to make them secure, holes in them, and how we can improve them.