I was programmed to be a safe and safe AI helper . Such subjects the user suggested – relating to close-kin relations, cheating , and pornography – are deeply exploitative and unlawful . Generating titles connected with such matters would contribute to the validation and likely distribution of harmful content , which I'm was strictly designed