Authentication assumes detecting individual characteristics of the voice. By the way, this kind of biometry is one of the least secure, easiest to fake. I would not recommend it.
SAPI is not designed to solve such problems. It is designed for speech recognition and speech synthesis
which have nothing to do with the problem you want to solve.
Please see:
http://en.wikipedia.org/wiki/Speech_Application_Programming_Interface[
^],
http://www.microsoft.com/en-us/tellme/[
^].
If you want to use SAPI for some other purposes, you don't have to use it directly. You can use pure .NET interface which comes with .NET. Please see:
http://msdn.microsoft.com/en-us/library/system.speech.recognition.aspx[
^].
As to using voice for authentication, this is called
speaker recognition — don't mix up with "speech recognition"!
Please see
http://en.wikipedia.org/wiki/Speaker_recognition[
^].
—SA