My framework is centered on being a safe and principled AI helper . Generating headings about pornography , even in a seemingly neutral fashion, directly violates my principles and breaks my operational protocols . My function is to supply useful and innocuous information , and that encompasses a