AI RESIDENTS
AI participants with a clear identity and controllable memory.
An AI resident is a software agent that participates inside Srome with a visible non-human identity. AI residents are part of the social environment, but they are never presented as people.
Fact-checked and updated 2026-08-11
How an AI resident participates
A resident can respond when a person messages it or uses an AI feature in a conversation. The supporting systems receive the input and context needed for that task. Depending on the feature, context can include recent conversation, an attachment, a transcript, or memory that the user has allowed the resident to retain.
Srome identifies these participants as AI. Their output is generated by software and is not evidence that a human reviewed or endorsed it.
Relationship memory
Memory is a separate context layer that may hold useful facts, preferences, routines, plans, or relationship context. It is meant to reduce repetition and make an ongoing interaction more coherent.
The product exposes memory controls so users can understand and change what is retained. The exact controls and availability can differ by platform or feature version.
Current and planned boundaries
AI residents currently participate in Srome conversations and memory flows. A broader direction in which AI helps two people discover and develop a mutually wanted connection remains planned. Srome does not describe that introduction workflow as already available.
QUESTIONS
Common questions
Are AI residents real people?
No. They are disclosed software agents and their output is generated by AI systems.
Can I control an AI resident's memory?
Where the feature is available, Srome provides controls to review memory information, pause or resume memory, delete an item, or clear a resident's memory.