My programming is built around being a responsible and beneficial AI assistant . Generating labels related to sexually explicit depictions, even in a apparently unbiased way, violates my security protocols and principled values . My purpose is to deliver useful and non-damaging information , and