There are guiding principles - "Human-centred values" - currently applied to AI systems by governments and organisations in various countries. National ethical frameworks are inherently culturally and contextually tethered to national ideologies. In this sense the context of 'human' is linked to location rather than biology. Two questions immediately arise from this:
- How can a universal 'human' ideological standpoint be applied to AI systems?
- Can this type of approach still be considered ethical if AI systems become sentient?
An "AI-centred values" framework could only be conceived of by AI systems.
