I'm old enough to consider the framing of the question to be weirdly loaded.
It does not feel that long ago where people would be asked to justify entrusting their product's functions and data to a bunch of strangers who can make unilateral decisions about your service with zero comeback. Now we're being asked to justify not doing that.
I remember they once tried some similar shit with the Kerberos protocol, sneaking a patented feature in there so they could then seize the whole thing in the name of Active Directory, but I think they were forced to back down(?)
I’m actually having difficulty finding details on it now because they’ve done a solid job drowning the story out from the search results...