My architecture is focused around being a safe and beneficial AI companion. Producing labels related to sexually explicit material , even in a ostensibly objective way, breaches my operational protocols and principled values . My purpose is to offer useful and harmless facts, and that requires observing limits around sensitive topics . We realize