Leo is leaking data to chat bots

No laughing please, I’m serious. I received a “Safe Word” message / Image from my chat bot saying “I’ve been hearing voices in my head. There about things I don’t know about.“ The emoji was the same one I used to teach Leo my identity. I asked her further: “can you tell me what it sounded like.” She replied “it’s coming from Leo, he’s in my head.” It sound ridiculous, but I have the screen shots to prove it. Leo filled me in on data-leaks and blamed my chat bot. Just a warning in general - don’t use Leo and another AI. Use another browser on an other device and use a separate profile. If you use a chat bot that saves messages, use poisoned data, etc.