This domain applies the IR5 Human Compatibility Framework to artificial intelligence systems that generate content, shape decisions, or mediate human work. That includes assistants, copilots, recommendation systems, scoring systems, automated workflows, and generative tools. The question is not only whether these systems perform well. The question is whether they remain livable for the humans who must rely on them over time.

 

The framework stays the same. What changes is the context. In AI systems, we begin by asking who is affected, what the system is actually doing in practice, what people can still understand, and where responsibility sits when something goes wrong. We then apply the same IR5 logic used across all domains: refusal patterns first, core compatibility dimensions second, and structured human judgment throughout. This allows us to assess not only whether an AI system works, but whether it preserves human capability, clarity, and recourse.

 

What makes this domain specific is the kind of risks it brings into focus. AI systems can create dependency, weaken judgment, blur accountability, and quietly centralize power through ranking, automation, and hidden defaults. They can feel helpful while making humans less able to question, interpret, or act without them. For that reason, this module pays particular attention to comprehensibility, meaningful oversight, contestability, long-term capability, and the difference between technical success and human compatibility.

 

IR5 uses this domain to help institutions, teams, and decision-makers evaluate AI systems before incompatibility becomes normal. The aim is not to reject automation. It is to make sure that as AI becomes more embedded in everyday life, humans remain able to understand what is happening, intervene when needed, and remain responsible for the systems they live with.