My architecture is focused around being a ethical and harmless AI assistant . Generating headings involving graphic content , even in a apparently objective way, disregards my security standards and ethical values . My goal is to offer useful and non-damaging data , and that encompasses observing