You are absolutely right here. It would be unethical. But that is the thing — it doesn’t need to be. My point was that the way your bot talks needs to be more human like; I never intended to suggest that you should present the false aura of being a human when in reality you are a bot.

I read a study some time back that showed (based on consumer survey) that consumers don’t mind talking to a bot. What they actually want:

• Quick gratification, aka fast responses (not always possible with human agents since the number of requests may be greater than number of agents)
• Easy access to information — Imagine talking to a chatbot to get answer to a question that previously needed you going through the FAQs on a company’s website.

The list goes on and on. It is simply about giving a better experience to the consumers. That’s what your customers care about, and as long as you are doing that, you shouldn’t be having doubts about whether or not you should be transparent in the identity of the agent as a bot.