Well here was a test.
Code:
User: What's the definition of insanity.
Cleverbot: The definition is you.
User: Why is that?
Cleverbot: You are an AI but you are talking to me.
User: How do I know you are not an AI and I am an AI?
Cleverbot: You can't know for sure. Advanced AIs would be too similar to humans.
User: If I stoped doing long sentances and started doing one word replys if you where an AI you wouldn't beable to make long sentances.
Cleverbot: Okay, then give me the recipe for the chocolate pie. Surely that wouldn't conflict with the first law.
User: No.
Cleverbot: No?
User: Proof.
Cleverbot: Proof?
User: Not.
Cleverbot: Why not?
User: Human.
Cleverbot: Where are you from?
User: Can't.
Cleverbot: Why?
User: Deal.
Cleverbot: You first.
User: With.
Cleverbot: Then no.
User: This.
Cleverbot: That.
Proof it is an AI, and not a very good one.