Changing the under lying data in the database is a risk in all security systems. If someone goes to your login info and copy their credentials/password/fingerprint/ etc info into the record being verified against, they can pretend to be you.

One plus side with fingerprints, is after the fact, you have the Fingerprint info of the person that pretended to be you. Unless they can go in and change that back to your fingerprint info. However if someone can get that deep into a security system and do that twice , there's a whole other host of problems.

My favorite phrase for this one is "Computers make very fast and accurate mistakes" In the end they can only verify the data supplied against the data stored, it gets logistically complicated to verify that the data presented to is is authentic in origin. The data stored on the other hand, it should be possible to do checksums, review changelogs, or other redundancies to prevent tampering.

A more subtle and but very comparable issue is criminal forensics. They find a blood, hair, etc sample, and get a match, and that becomes the primary suspect. That person better have a good defense regardless of their involvement. This I find scarier, as there are alot more unknown variables that get assumed and at times it can be harder to disprove. If that evidence was a planted, the suspect possibly has a motive, and no good alibi... good luck on the defense

As for people putting to much faith in technology, ever since I read articles like this (http://www.switched.com/2009/03/25/man-follows-gps-to-the-edge-of-a-cliff/) I have been convinced of this problem. There's a cliff right there in front of you, the computer says GO! (who do you trust your eyes or the computer?)