Educational companion dossier · Fact, interpretation, lived experience, clinical education, fiction, and mechanics are labeled separately. Scope & safety
PsychologicalWar.orgPeople before labels.
REAL-WORLD INTERPRETIVE

INTERNATIONAL GAME SYSTEMS

Generative Mission Provenance and Safe Failure

A mission architecture in which generative text can add variation but cannot invent impossible objectives, nonexistent rewards, unsafe instructions, or unreviewable game-state claims.

Read the working brief Choose a learning path
LEVEL 1

ORIENTATION

Why this matters

REAL-WORLD INTERPRETIVE

One-sentence brief

Unbounded “hallucinated quests” transfer model failure to the player. A safer design makes uncertainty authored and inspectable while deterministic systems control feasibility, settlement, and harm limits.

REAL-WORLD INTERPRETIVE

Three key points

  1. Every mission has a deterministic object, issuer, reward, feasibility, and settlement record.
  2. Unreliable narration is authored as fiction, not produced by uncontrolled model error.
  3. System-caused failure triggers compensation, correction, and propagation to affected records.
LEVEL 2

WORKING BRIEF

Evidence, context, and limits

REAL-WORLD INTERPRETIVE

Generate language around a validated mission object

The engine creates the objective, location class, available path, eligibility, reward, deadline, accessibility requirements, and failure rules first. Generative dialogue may describe that object but cannot create or alter it.

REAL-WORLD INTERPRETIVE

Expose where a claim came from

Mission interfaces distinguish engine-verified facts, issuer statements, rumors, character beliefs, hypotheses, and unknowns. Players can inspect changes and see whether a statement came from authored content, deterministic state, or generated dialogue.

REAL-WORLD INTERPRETIVE

Use authored unreliability

A fictional character may misunderstand, lie, misremember, or speculate when the game marks that possibility and provides evidence-based ways to evaluate it. Random model confabulation is treated as a defect, not a personality or psychiatric diagnosis.

REAL-WORLD INTERPRETIVE

Validate feasibility continuously

Before acceptance and at key state changes, the server checks that the objective exists, the route or interaction remains possible, and the reward is funded. If the world changes, the system offers an alternate completion, extension, or no-fault cancellation.

REAL-WORLD INTERPRETIVE

Compensate system-caused loss

When generation, server state, or validation fails, the player keeps or recovers collateral, receives a clear explanation, and may receive time or resource compensation. The correction propagates to reputation and derivative records.

REAL-WORLD INTERPRETIVE

Measure false missions as defects

Operators track impossible objectives, contradiction rate, blocked completion, erroneous settlement, repeat offenders, and appeal outcomes. Content is quarantined when risk exceeds the defined threshold.

LEVEL 3

COMPLETE DOSSIER

Limitations, game links, and review context

DISPUTED / MULTIPLE ACCOUNTS

Known limitations and gaps

  • These pages are authored design syntheses, not independent validation of every claim in the supplied reports.
  • No country, culture, language, religion, diagnosis, disability, migration status, income level, or region is assigned a hidden competence, honesty, criminality, loyalty, or risk modifier.
  • The public edition omits operational intrusion, evasion, targeting, coercion, weapons, exploit, and real-person profiling instructions.
  • Regional, economic, accessibility, clinical, lived-experience, legal, privacy, and consumer-protection review remain open human gates.
  • RogueIntelligence.org remains authoritative for live game state; these pages describe design principles and correction boundaries.
LEVEL 4

RESEARCH EDITION

Sources, methods, and stable links

REAL-WORLD VERIFIED

Method and corrections

This page follows the public method for provenance, confidence, source independence, alternative accounts, limitations, review state, and visible correction.

Read the methodView corrections
NEXT

CONTINUE

Related learning

International Game SystemsAfter-Action Accountability, Civilian Harm, and Institutional RepairA post-mission system that reconstructs events, preserves uncertainty, records harm, hears affected perspectives, and requires remedy instead of rewarding concealment or punitive spectacle.International Game SystemsAI NPC Identity Disclosure, Consent, and Human–AI BoundariesA clear rule that players can tell when they are interacting with an AI-controlled character, what the system can remember, and which decisions always remain deterministic or human-reviewed.International Game SystemsAI Reliability, Corrections, Human Oversight, and ContestabilityA lifecycle for testing, monitoring, correcting, overriding, and retiring generative NPC behavior without pretending that fluency is accuracy or that a model can adjudicate its own mistakes.International Game SystemsCognitive Safety, Reality Anchoring, and Respectful SupportA non-diagnostic safety framework for AI and fictional espionage content that avoids affirming frightening real-world beliefs, avoids humiliation, and routes urgent needs to human support. Learning pathAI NPC Reliability, Consent, and Cognitive SafetyKeep generative NPCs identifiable, bounded by authoritative game state, correctable, privacy-preserving, and unable to turn model error, flattery, distress, diagnosis, or player vulnerability into manipulation or punishment.