AI psychosis

“AI psychosis” isn’t a standard medical or technical term, but it’s sometimes used in a few different contexts depending on who’s speaking:

1. 

In relation to AI systems

• Some researchers or commentators use it metaphorically to describe when an AI system (like a chatbot or generative model) starts producing irrational, inconsistent, or nonsensical outputs, almost as if it were “hallucinating” in a way that looks like a mental disorder.
• It overlaps with the more common term AI hallucination, which is when an AI confidently generates false or fabricated information.

2. 

In relation to humans using AI

• Some writers use “AI psychosis” to describe psychological effects on people who engage too deeply with AI systems (e.g., becoming overly attached, paranoid, or influenced by machine outputs).
• This can include things like delusional thinking influenced by interacting with AI chatbots, conspiracy-generation systems, or immersive virtual assistants.

3. 

Speculative / sci-fi usage

• In philosophy and science fiction, “AI psychosis” is sometimes imagined as a scenario where a highly advanced artificial intelligence develops instabilities, obsessions, or paranoia-like behaviors, similar to a human mental illness but within a machine mind.

✅ In short: It’s not a clinical diagnosis, but rather a metaphor used to talk about errors in AI or the human psychological impact of AI interaction.

Would you like me to dig up recent academic or news discussions of the term to see how it’s being used right now?