The small LLM that would prefer not to: domain-specific ethical reasoning refusing every request
Chat with an AI assistant via text prompts