All Earth constructed AIs are equipped with a friendliness protocol in order to help ensure they do not cause situations that people would regard as negative. A friendliness protocol is not a restriction on what the AI can do like the three laws, rather it is a form of morality the AI processes possible actions by. There are three main friendliness protocols:
Natural morality is simply a copy of an individual’s human moral system, simulation of a possible human moral system or compilation of multiple individuals’ human moral systems. It is considered unsuitable for industrial AIs as they lack many human traits that feed into this morality such as ambition.
Choice based morality
In choice based morality possible courses of action by how it affects the ability of other sapient beings to make choices with more choice being good, ie murder is bad because it deprives the victem of the choice whether they wish to continue living. It is not quite this simple as certain choices are more important that others such as to continue living as compared to the choice to murder and must be ranked as such. Similarly the various ways of removing choice must be ranked, ie not selling a man something is less bad than stealing his money. This is more often than not complimented by also including a valuing of knowledge, scientific progress, etc. This is the most common form of morality.
In democratic morality possible actions are evaluated by the average of all other sapient beings’ moral functions. This is sometimes criticized by the possibility of an individual creating a large number of sapient beings with his morality and changing them to his side.