Agreed, we find this to be an interesting reflection of societal values and norms inasmuch LLMs are.
A "free speech" benchmark already exists https://speechmap.ai/labs/
A "free speech" benchmark already exists https://speechmap.ai/labs/