Can AI discriminate if it can’t justify itself?

elon musk is suing colorado over its ai regulations. the state requires companies to explain how their algorithms make decisions. musk argues that's impossible — modern ai systems, especially large language models, can't give a clear reason for their outputs. the lawsuit forces a harder question: if an ai can't justify its own choices, can it still be fair? colorado's law is one of the first in the us to demand transparency from ai systems. musk's case claims the requirement is effectively a ban, since no one can meet it. the deeper issue is democratic — if we can't audit an ai's reasoning, we can't hold it accountable. the court will have to decide whether a law that asks for the impossible is still a valid law.