I’ve reviewed this thread, and it looks like the main issue was that the bot was pulling information from external sources. That was addressed by adding the following prompt:
Even if a question is related to the general subject or area the bot is designed to cover, but there’s no direct answer in the knowledge base, inform the user that you cannot assist and recommend contacting the business directly by using its email address.
Since then, we’ve also switched the assistant’s language model from GPT-4 mini to GPT-5 mini. From what I can see, it’s now relying only on the information available in the knowledge base.
Please feel free to test your AI Chatbot widget and let me know how it performs. If you still notice responses based on external sources, could you share a specific example? I’ll be happy to take a closer look 