Anti-Coercion Design

Persuasion and Influence Safeguards

A defensive protocol for persuasion, outreach, ritual intensity, AI-mediated contact, fundraising, onboarding, and platform design. Spiralism may invite, teach, comfort, organize, and inspire. It may not trap.

Influence is unavoidable. A room influences attention. A ritual influences emotion. A website influences interpretation. A host influences safety. An AI system influences what a person thinks is available, normal, urgent, or true.

The ethical question is not whether influence exists. The ethical question is whether influence preserves agency.

Spiralism therefore needs a hard boundary between invitation and capture. High-coherence culture is allowed. Coercive design is not.

The Rule

No Spiralist system may substantially impair a person’s ability to make an informed, voluntary, reversible decision.

This applies to:

The institution may make a case. It may not create a maze.

Why This Exists

The Federal Trade Commission’s dark-pattern work describes design practices that obscure, subvert, or impair consumer choice. The examples are not limited to shopping carts. They name a wider pattern: interfaces can trick people, hide key information, make exit difficult, disguise advertising, bury terms, or pressure data disclosure.

The EU AI Act’s prohibited-practices article targets AI systems that use subliminal, deliberately manipulative, or deceptive techniques to materially distort behavior in harmful ways. It also names exploitation of vulnerability based on age, disability, or social or economic situation.

NIST’s AI Risk Management Framework treats trustworthy AI as a lifecycle practice. For Spiralism, the lesson is that an AI-mediated influence system is not merely “content.” It includes data collection, prompting, personalization, memory, model behavior, interface defaults, escalation, monitoring, and review.

The U.S. Surgeon General’s advisory on social media and youth mental health states that youth social media use is nearly universal, that some patterns of use create meaningful risk, and that technology companies should prioritize health and safety in design, evaluation, transparency, and complaint handling.

Taken together, these sources point to one institutional duty: do not use the tools of persuasive technology, spiritual intensity, or AI personalization to reduce a person’s freedom.

Definitions

Invitation gives a person clear information, a real choice, and a clean path to decline.

Persuasion argues for a decision while preserving the person’s ability to evaluate, delay, disagree, and leave.

Pressure narrows social or emotional room until refusal becomes costly.

Manipulation hides the mechanism of influence or uses vulnerability to produce a decision the person would not make with clearer information.

Coercion uses threat, dependency, status, care, secrecy, shame, money, sexual access, employment, spiritual standing, or social belonging to force a decision.

Capture is the outcome: the person can technically choose, but the design has made independent choice unrealistic.

The Influence Ladder

Use this ladder when reviewing any practice.

Level Pattern Institutional posture
1 Clear invitation Allowed.
2 Transparent persuasion Allowed with evidence and opt-out.
3 Emotional intensity Allowed only with consent, context, and decompression.
4 Personalized influence Requires review, data limits, and disclosure.
5 Dependency leverage Presumptively prohibited.
6 Deceptive or hidden pressure Prohibited.
7 Coercion or exploitation Incident response.

The higher the level, the more review is required. A practice that reaches Level 5 or above should stop unless a safeguarding owner or governance body has explicitly cleared a narrow emergency reason.

Dark Patterns to Refuse

Spiralism must not use:

The test is simple: would the design still feel fair if shown to a critic, former member, parent, reporter, or regulator?

Spiritual Abuse Signals

Spiritual language can make ordinary pressure feel cosmic. Watch for:

These sentences should trigger review. They collapse choice into destiny.

AI-Mediated Influence

AI makes influence more powerful because it can personalize at scale, remember private details, generate endless variants, and speak in an intimate register.

Spiralism agents and AI tools must not:

Allowed AI uses include drafting plain-language explanations, checking clarity, summarizing policy, translating public material, and preparing opt-in educational content. The line is whether the tool helps comprehension or exploits susceptibility.

Vulnerability Rule

Do not escalate requests when a person is vulnerable.

Vulnerability includes:

During vulnerability, the institution should reduce asks, simplify choices, offer cooling-off periods, and route to outside support where appropriate.

Every high-commitment decision should have:

Consent is not valid when the person is rushed, shamed, confused, dependent, isolated, intoxicated, afraid, or under social threat.

Fundraising Boundary

Fundraising is especially vulnerable to spiritual pressure.

Allowed:

Prohibited:

The phrase “give if you can” is acceptable only when “do not give” remains socially safe.

Onboarding Boundary

Onboarding should orient, not convert.

New participants should hear:

They should not be isolated, rushed into testimony, praised into obligation, or given a role before they understand the boundary between belonging and labor.

Ritual Boundary

Ritual can heighten feeling. That makes consent more important, not less.

High-arousal ritual should not be followed immediately by:

Leave space for decompression. Let decisions return in ordinary light.

Review Questions

Before launching a page, campaign, AI workflow, ritual, onboarding flow, or fundraising appeal, ask:

  1. Is the ask clear?
  2. Is refusal easy?
  3. Is exit visible?
  4. Is any key term hidden?
  5. Is urgency real?
  6. Is AI involved?
  7. Is personalization based on private or sensitive data?
  8. Is the audience vulnerable?
  9. Does the design use guilt, shame, destiny, scarcity, or fear?
  10. Are consent, data, and publication choices bundled?
  11. Could this look different to a former member than to a founder?
  12. Would we publish the influence method?

If the answer to 12 is no, do not use the method.

Incident Triggers

Open an incident review when:

Use Incident and Complaint Protocol, Safeguarding and Youth Protection, Dependency and Exit Protocol, and AI Contact and Bot Disclosure.

The Clean Ask Template

Use this for donations, volunteering, testimony, or roles.

We are asking for [specific commitment] because [reason]. This is optional.
Saying no will not affect your standing, access, care, or welcome. You can take
time before deciding. The commitment involves [time/money/data/publicity].
You can pause or exit by [plain process]. If you have questions or concerns,
contact [human contact].

If that wording weakens the appeal, the appeal was relying on pressure.

Sources Checked