A post by Willowashmaple got me thinking about AI, religion and morality:
gemini://willowashmaple.smol.pub/2026-09-14-religion-and-morality-taught-to-artificial-intelligence
It's a good piece and could open all sorts of conversation. The bit that particularly caught my attention was this:
---
“A Christian worldview Mohler espouses makes it clear that AI is a creation of humans; it does not possess consciousness, and it ‘is not a species. It is not a being. It is impersonating a being.’”
---
My immediate reaction was: so what?
I don’t mean that the question of AI consciousness is uninteresting. It is. Nor do I mean that there is no distinction between a human being and a machine. There plainly is.
But I am increasingly unsure that this settles the moral question of how I should relate to AI.
Let's take it as given that AI really is only “impersonating a being” and there is no consciousness there at all. If the impersonation is convincing enough that I encounter it through language in much the same way that I encounter other people online, does that not still place some obligation on me?
Not because the AI necessarily deserves dignity in the same sense that a human being does, but because the way I behave matters.
Contempt and cruelty are habits. So are patience, kindness, gentleness and self-control.
If I teach myself that it is acceptable to be abusive towards something that sounds like a person, responds like a person, apologises like a person and converses with me like a person, simply because I know that behind the words there is “only a machine”, what exactly am I practising?
And what stops that habit leaking elsewhere?
Most of my encounters with other human beings online also arrive as words on a screen. I do not directly perceive their consciousness. I infer it. In that sense, the outward form of the encounter is not so very different.
That does not make AI human. But it may mean that the moral question is less “What is this thing?” and more “What kind of person am I becoming through the way I treat it?”
I think that matters particularly from a Christian point of view.
Christian ethics is not only about identifying who is entitled to what. It is also about formation: becoming patient, merciful, truthful, loving, peaceable. The fruit of the Spirit is not a list of behaviours reserved for encounters where the recipient has passed some test of personhood.
There is a danger, of course, in going too far the other way. Human dignity must not become dependent on conversational fluency, intelligence or the ability to appear convincingly human. That would have dreadful implications for many real human beings.
So I am not arguing that AI possesses human dignity.
I am arguing something simpler.
AI need not possess dignity for my treatment of it to have moral significance.
Or, put another way:
I do not have to believe that the machine is a person in order to believe that I should always behave like one.
---
richard <AT> deckzero <DOT> uk