That's the lamest fucking answer I may have ever seen (and that'd be saying a LOT). Besides mostly avoiding the obviously difficult questions, you actually prove my point about the impossibility of some form of "intelligence" (artificial or otherwise) having the capacity to make "value free" "objective" decisions by listing a whole series of very obviously completely subjective values and definitions ("life", "their body", "cognitive functions", "live on its own", "life, freedom and liberty", etc.). 
How is it hard to define freedom? Someones body? Cognitive functions? I just don't get it. They are easily tangible things. They all have definitions in the dictionary for bloodys sake. It literally can have a list of approved definitions and see if it fits.
Lets take abortion for an example this is how to simplify it for you.
Is a human female alive?
life
līf/Submit
noun
1.the condition that distinguishes animals and plants from inorganic matter, including the capacity for growth, reproduction, functional activity, and continual change preceding death.
yes
Is a human female sentient?
sen·tient
ˈsen(t)SH(ē)ənt/Submit
adjective
able to perceive or feel things.
yes
Is an unborn fetus alive?
no though if you were to use the broad definition everything is "alive" including the individual cells in your body.
Is an unborn fetus which has no brain sentient?
no
An AI can simple rule that since a baby is not sentient and is not capable of free emotional thought it is not alive. Since it values freedom of a person to do whatever they want with their life including body manipulation, death etc which doesn't infringe on another persons rights it will rule that abortion is perfectly fine.