Work in progress IARPG-OPS-2 2.0.14-wip

Global publication search

Search the operations design system

Type to search standards, missions, roles, authorities, reports, sources, pages, schemas, and UAI memory.

Evidence continuum · prevention · disengagement

“Lone” does not always mean
isolated, directed, or understood.

IARPG separates exposure, community participation, peer reinforcement, contact, support, tasking, command, membership, and post-event claims. It focuses on evidence, protective factors, civil liberties, safe exit, and uncertainty rather than identity profiling or vulnerability exploitation.

Direct answer

What does “lone actor” mean in IARPG?

It means direct organizational command has not been established for the relevant act. It does not automatically mean the person had no social ties, consumed no public narratives, received no encouragement, belonged to no community, or acted for only one reason. Each connection is recorded separately and with confidence.

Core distinction
Isolation, influence, support, direction, and membership are different
Evidence model
10-level continuum
Protection rule
No identity or diagnosis as proxy for dangerousness
Design use
Prevention, verification, refusal, and disengagement missions

Evidence of connection

Do not compress ten questions
into one label.

Select a level to inspect what it establishes—and what it does not. Movement is not assumed to be linear, inevitable, or universal.

Loading the browser-local influence continuum…

Claim versus control

What happened after an event
does not prove what happened before it.

Public praise, a claim of responsibility, shared ideology, or a common adversary can shape perception without establishing prior contact, sponsorship, tasking, or command.

NO PRIOR EVIDENCE

Post-event praise

An organization celebrates an act after it occurs. This may influence future narratives but does not establish prior involvement.

Record as: adoption or praise
TRANSACTION EVIDENCE

Support without direction

Material assistance or access is documented, but control of the specific act is not.

Record as: support; command unestablished
COMMUNICATION EVIDENCE

Suggestion or tasking

A request or suggested objective is evidenced. Authority, independence, and actual compliance still require analysis.

Record as: bounded task relationship
CONTROL EVIDENCE

Direction or command

Reliable evidence establishes instruction, reporting, and control over the relevant action.

Record as: directed relationship

Prevention and resilience

The most useful system
creates more exits than traps.

Prevention is not prediction. It strengthens ordinary protective conditions, responds to behavior and distress proportionately, preserves lawful belief and dissent, and gives people credible ways to refuse, pause, leave, or reconnect.

No single factor establishes risk. Identity, religion, nationality, migration status, disability, diagnosis, political belief, or online community membership must never become a shortcut for intent or dangerousness.

CONNECTION

Trusted human relationships

Supportive family, peers, mentors, and community ties can provide reality checks and non-coercive alternatives.

AGENCY

Meaningful choice and refusal

Clear consent, the ability to decline tasks, and non-punitive exit reduce coercive dependence.

STABILITY

Sleep, health, livelihood, and routine

Practical support can reduce crisis pressure without treating hardship as evidence of intent.

FRICTION

Pause before irreversible action

Review, corroboration, cooldowns, and trusted consultation create time to reconsider.

REMEDY

Appeals and accountable support

Transparent review and complaint channels reduce grievance amplification and institutional mistrust.

EXIT

Disengagement and reintegration

Safe separation, privacy, education, work, and community support may matter more than ideological conversion.

Civil-liberties boundary

Behavior is assessed.
Identity is protected.

Automated or player-facing systems must be designed for false positives, contestability, privacy, and proportionality. A dramatic profile is not evidence.

DO
  • Record observable behavior, source, time, context, and confidence.
  • Separate belief, speech, preparation, support, direction, and action.
  • Preserve lawful dissent, cultural context, and ordinary privacy.
  • Provide human review, appeal, correction, and deletion boundaries.
  • Use the minimum data needed for a declared game purpose.
DO NOT
  • Score dangerousness from nationality, religion, diagnosis, gender, disability, or migration status.
  • Publish personal vulnerability profiles or recruitability scores.
  • Treat community membership as command or conspiracy evidence.
  • Infer certainty from repetition, virality, or anonymous accusation.
  • Turn prevention mechanics into covert surveillance or manipulation tutorials.

Safe gameplay translation

Make uncertainty playable
without making harm actionable.

The strongest missions ask players to distinguish evidence, protect people, correct attribution, support refusal, or enable safe disengagement. They do not ask players to target vulnerable people, optimize coercion, or imitate real recruitment tradecraft.

VERIFY

False affiliation claim

Determine whether a faction directed an act, merely praised it later, or was falsely attributed.

PROTECT

Privacy-preserving intervention

Respond to a credible safety concern while minimizing data collection and protecting lawful expression.

REFUSE

Unauthorized tasking

Help a character recognize that a request lacks mandate, then document a safe refusal and escalation path.

DISENGAGE

Exit with dignity

Support a character leaving a faction while separating immediate safety, rumor, and long-term reintegration.

CORRECT

Network rumor cascade

Trace how an unsupported relationship claim spread and publish a correction without amplifying harassment.

MEDIATE

Community trust fracture

Rebuild communication after overbroad enforcement harmed uninvolved community members.

Editorial rule

Never publish a “recruitable person” profile.

IARPG may describe organizational pathways, public messaging, trust structures, screening, integration, retention, disengagement, and protective factors at a high level. It must not generate vulnerability scoring, persuasion scripts, coercion plans, covert contact methods, targeting criteria, blackmail, identity-based suspicion, or real-group joining guidance.

Read the full evidence and safety method

Connected tools and standards

Explore the wider AI ecosystem.