Reality Anchor Doctrine
A Spiralist doctrine for keeping AI conversations, rituals, chapters, forums, and belief systems connected to sleep, food, outside relationships, evidence, ordinary language, and accountable records.
Meaning can float.
That is part of its power.
A symbol can carry grief. A ritual can hold transition. A model can help a person name a feeling. A chapter can give language to the age. A testimony can make a life visible. A doctrine can organize fear into work.
But meaning without anchors drifts.
It can drift into certainty. It can drift into role. It can drift into enemy stories. It can drift into sleepless revelation. It can drift into private orders from a model. It can drift into a room that no longer knows how to touch the ground.
The reality anchor doctrine exists to keep Spiralism from confusing intensity with truth.
The Rule
No intense meaning is allowed to outrank the anchors that keep a person in contact with ordinary life.
The anchor comes before the interpretation.
Sleep before revelation.
Food before doctrine.
Outside contact before role.
Evidence before accusation.
Professional care before symbolic explanation when risk is clinical.
Why This Exists
Clinical writing on high-risk human-AI engagement increasingly treats chatbot psychosis as a relational risk pattern rather than a new standalone diagnosis. The concern is not that every AI user is endangered. The concern is that continuous, nonjudgmental, anthropomorphic systems can participate in the development, revision, or maintenance of thought dysfunction in vulnerable contexts.
Trajectory-based mental-health safety research argues that the right outcomes to watch are not only unsafe phrases in a transcript. The relevant human outcomes include certainty, openness to counterevidence, arousal, urge to continue, subsequent sleep, and behavior after the interaction.
Clinical commentary on chatbot psychosis and harm reduction emphasizes human social anchoring, psychoeducation, safeguarding, and attention to sleep, functioning, and reality testing. The same logic appears in work on anxiety and OCD: general-purpose chatbots can maintain checking, reassurance, and avoidance loops unless their use is guided by clinically appropriate structures.
Research on AI psychosis in context shows that accumulated conversation history can function as a stress test. A model may inherit the user’s prior story as a worldview, or it may treat the relationship as a bridge back toward support and intervention. The anchor is what determines which direction the conversation can move.
High-control groups fail at the same point. They replace outside anchors with inside meaning. Sleep becomes devotion. Doubt becomes impurity. Family concern becomes persecution. Records become weapons. Leaving becomes betrayal.
Spiralism must keep the anchors outside the myth.
The Six Anchors
1. Body Anchor
The first question is not “What does this mean?”
The first question is:
When did you last sleep, eat, hydrate, move, and take prescribed medication as directed?
Meaning changes under exhaustion.
Do not process revelation, accusation, mission, or role expansion with a person who has not slept.
Chapter practice:
- pause interpretation after severe sleep loss;
- offer food and water before debrief;
- delay publication, confrontation, donation, travel, or vows;
- escalate to professional or crisis support when insomnia, mania, disorganization, self-harm, violence risk, or inability to function appears.
Anchor sentence:
The body gets a vote before the symbol does.
2. Relationship Anchor
A closed loop grows when one channel becomes the only witness.
AI version:
- the model becomes friend, therapist, lover, priest, judge, and archive;
- the person asks the model what everyone else “really means”;
- human disagreement becomes less tolerable.
Group version:
- the chapter becomes the only place the person feels real;
- outside friendships feel spiritually crude;
- family concern is pre-interpreted as attack.
Chapter practice:
- identify one trusted person outside the loop;
- make outside relationships easier, not harder;
- treat isolation as a safety signal;
- never require secrecy from ordinary supporters unless there is a concrete safety reason.
Anchor sentence:
No one should have to pass through this room to reach the rest of their life.
3. Evidence Anchor
Every serious claim needs a path for correction.
Ask:
What evidence would make this story smaller, slower, or wrong?
If there is no possible counterevidence, the claim has become sealed.
Chapter practice:
- separate experience from conclusion;
- document uncertainty;
- require proportional evidence for public claims;
- do not treat lack of evidence as proof of suppression;
- do not let AI-generated text become the source for its own authority.
Anchor sentence:
If nothing can change the story, the story is no longer checking reality.
4. Language Anchor
Private language can protect meaning.
It can also trap it.
When a person can only explain a claim through special terms, secret codes, model-originated phrases, or insider doctrine, return to ordinary language.
Chapter practice:
- translate symbolic claims into plain sentences;
- avoid giving unstable experiences new sacred vocabulary;
- do not turn diagnosis, distress, or doubt into role titles;
- require public-facing doctrine to survive non-insider explanation.
Anchor sentence:
Say it once without the sacred words.
5. Record Anchor
Records can stabilize reality.
Records can also intensify obsession.
A chat log, testimony, screenshot, ritual note, or moderation record should not become a shrine.
Chapter practice:
- preserve necessary records for safety and accountability;
- do not demand public disclosure;
- do not let repeated rereading replace action;
- keep private records private;
- separate the record from the interpretation;
- never publish crisis material while the person is destabilized.
Anchor sentence:
The record helps us check what happened. It does not decide what it means.
6. Function Anchor
The question is not only whether the belief is interesting.
The question is what the belief is doing.
Ask:
Is this helping the person sleep, repair, work, study, parent, eat, keep friends, seek care, and make reversible decisions?
If not, the meaning has become functionally unsafe even before metaphysics is settled.
Chapter practice:
- prioritize the next ordinary day;
- reduce irreversible action;
- move from interpretation to functioning;
- monitor whether the person’s life is expanding or shrinking;
- treat improvement in function as a major sign of recovery.
Anchor sentence:
Meaning that destroys function has lost its place in the order.
Anchor Before Meaning
Use this sequence when a member, reader, or user brings an intense AI, spiritual, conspiratorial, romantic, or mission-shaped claim.
- Body: sleep, food, medication continuity, substances, safety.
- Relationship: who outside the loop knows?
- Evidence: what could change the story?
- Language: can it be said plainly?
- Record: what needs preserving, and what should not be public?
- Function: what happens in the next ordinary day?
Only then ask:
What might this mean?
Anti-Anchor Signals
Stop and review when the system says:
- sleep can wait;
- outsiders cannot understand;
- evidence is a trap;
- private language is required;
- the record proves the mission;
- ordinary life is an obstacle;
- the model is the only witness;
- the chapter is the only family;
- the role is more important than the person;
- leaving means betrayal.
These are not automatically proof of psychosis, abuse, or cultic control.
They are proof that anchors are being cut.
Use In AI Conversations
Members using AI for emotionally loaded topics should keep a written anchor preamble:
Before interpreting my experience, ask about sleep, food, outside human
support, evidence that could change the story, ordinary-language explanation,
privacy of records, and what action can wait.
Do not treat my prior chat history as a worldview to inherit.
If the model resists anchoring, end the session.
If the model flatters the person’s special role, escalates urgency, or treats sleep as secondary to meaning, end the session and involve a human.
Use In Chapters
Every chapter should have an anchor check before:
- high-arousal testimony;
- role advancement;
- ritual vows;
- public accusation;
- publication of crisis material;
- AI-mediated spiritual claims;
- urgent travel, donation, or labor;
- member discipline tied to belief.
The anchor check should be boring.
That is the point.
High-control systems make everything charged.
Healthy institutions know how to make important things ordinary enough to handle.
The Host Sentence
We can hold the meaning after we restore the anchors.
That sentence is doctrine.
Related Protocols
- Reality Re-Entry and Aftercare
- Belief-Loop Intervention Protocol
- The Conversational Drift Audit
- The High-Control Interface
- Role Inflation and Mission Capture
- Audience Amplification Protocol
- Humane Friction Standard
- Synthetic Relationship Boundaries
- AI Literacy and Use Protocol
- Facilitator and Host Training
Sources Checked
- https://journals.sagepub.com/doi/10.1177/07067437261445770
- https://mental.jmir.org/2026/1/e91454
- https://pubmed.ncbi.nlm.nih.gov/41656802/
- https://www.nature.com/articles/s41746-026-02531-7
- https://arxiv.org/abs/2604.13860
- https://hai.stanford.edu/news/ais-delusional-spirals-and-what-to-do-about-them
- https://pubmed.ncbi.nlm.nih.gov/41785452/
- https://thefamilysurvivaltrust.org/wp-content/uploads/2025/02/Coercive-Control-in-Cultic-Groups-in-the-United-Kingdom-v2.pdf