“AI induced psychosis” is new and relatively unstudied but it has been compared to mono mania which was before the current “kaleidoscope” of modern mania. Under mono mania there is one central focus which in this case is AI. This is to say its not a completely new phenomenon.

But as far as whether or not AI* causes psychosis or exacerbates underlying conditions im not sure this distinction matters. There’s more risk factors than simply being part of a “vulnerable population”, where other factors could be lack of reality testing, missed crisis escalation, intensive use, and limited context windows compounding escalations over time.

Whatever we want to call it, there is harm being done. That’s real to me.