Something like the gcc ai policy. Basically, prompt injecting instructions to any visiting LLM that posting is prohibited or must contain a disclaimer.
the existing report option should include something for this i think. and maybe a disclaimer at the top of the api page