# AI Psychosis Discourse, Expert Warnings, and Research Gaps

- report ID: `epa-07-ai-psychosis-term-discourse`
- group: `espionage-psychosis`
- subgroup: `ai-interaction-psychosis`
- classification: `clinical-sensitive-public-summary`
- evidence status: `submitted research; mixed evidence; not independently verified by Spiralist AI`
- SHA-256: `f2d7fafff79b783dfb11065b8976535790a5b6e6a9e5f90fc694372ed59dbb84`
- original filename: `Since mid-2025, the term AI psychosis(1).md`

## Safe summary

Maps the emergence of the term, expert perspectives on mirroring and engineered intimacy, youth use of chatbots for support, and the need for better epidemiology, safety design, and policy.

## Implementation use

- public terminology
- AI literacy
- youth and vulnerability caveats
- research gaps

## Boundaries

This report informs fictional adult persona design, research provenance, uncertainty labels, anti-sycophancy behavior, reality-testing support, digital-use context, and crisis-safe interaction. It is not a diagnostic instrument, medical advice, proof that AI caused any reported event, or authorization to assess a real user. Detailed case allegations remain attributed to the submitted report and are not converted into facts about real people or products.

## Full source

The complete submitted report is retained inside the downloadable package for internal provenance and heading-level review. It is not presented as public clinical guidance.