Chatbot Confessions: Should We Spill the Robo-Beans?

Title: Should AI Chatbots in Customer Service Disclose Their Identity?


I am amazed by the capabilities of ChatGPT-3.5-Turbo and GPT-4. By integrating these AI models into our workflows through API calls and other Python scripts, we have significantly improved our tour company’s customer service helpdesk.

As GPT-4’s responses now rival and even surpass human interactions, we are facing an ethical question: should our chatbot proactively disclose that it is an “AI Agent” or should it continue answering questions without revealing its identity unless it cannot provide an answer?

We are curious to hear your instinctive reactions and thoughts on whether chatbots should declare upfront that users are interacting with a bot. Our concern is that disclosing the chatbot’s identity might make the interaction feel impersonal, potentially causing some users to disengage from a source of valuable information. However, we also want to ensure our clients don’t feel misled.

Are there any insights, opinions, or experiences from the group?

PS>>> GPT 4 improved the structure of this question :))))))

1 Like