“This content may violate our policies” has crippled my ability to use GPT-4o. I use clinical terminology to describe symptoms. Out of context they may sound as violation, but they ARE NOT. The terminology I use is always present in the DSM-5
Can I get some help?
Are you diagnosing people with ChatGPT?
No. I enter MY diagnoses as part of the prompt when I am asking for treatment planning. I just save time with AI organizing my thoughts. Where I get issues is when I ask summaries of lengthy discombobulated family assessments from Child Protective Services. As one can imagine, they frequently refer to unsavory behaviors by parents
I’m getting similar messages repeatedly. It seems highly correlated to when I ask for serious revisions to python scripts that I created. Moreover, it strangely occurs when using o1-mini and o1-preview. I enter the same verbatim prompts or series of prompts in vanilla gpt4o and it at least responds.
I’ve posted elsewhere regarding this and asked for a way to “dispute” these or a system-wide change…keep me posted if you find a solution please.