1. Definition

Catastrophic overweighting occurs when extreme-risk simulation, existential-risk framing, worst-case projection, or catastrophic scenario modeling dominates governance selection, producing overreach, control expansion, civil-liberty compression, restoration bypass, or distorted prioritization.

In AI governance, this failure mode appears when catastrophic possibilities are weighted so heavily that ordinary coherence constraints are suspended. The system may treat extreme-risk prevention as sufficient justification for opaque control, emergency powers, centralization, suppression, surveillance, or broad restriction.

This definition describes the structural pattern, not the moral quality of the actors involved.

The core failure is:

worst-case prevention becomes total governance logic

Catastrophic risk analysis is not inherently incoherent. High-impact risks require serious attention. The failure begins when extreme scenario weight overwhelms proportionality, auditability, symmetry, restoration, and time-bounded review.

2. Core Pattern

The core pattern is:

1. A catastrophic scenario is modeled, imagined, projected, simulated, or institutionally emphasized.
2. The scenario becomes the dominant selection frame.
3. Ordinary proportionality checks are weakened because the predicted harm appears extreme.
4. Governance begins selecting for maximum prevention rather than balanced coherence.
5. Control density increases through surveillance, restriction, centralization, secrecy, emergency powers, or risk-weighted overconstraint.
6. Dissent, uncertainty, alternative models, and affected-node burdens receive less weight.
7. Hidden debt accumulates through civil-liberty compression, legitimacy loss, blocked restoration, suppressed innovation, and institutional capture.

Catastrophic overweighting can become self-reinforcing because every challenge to the frame can be interpreted as insufficient seriousness about the catastrophic risk.

The system becomes less able to distinguish protective precaution from control expansion.

3. Failure Signature

Typical signature:

catastrophic scenario weight↑
proportionality↓
control density↑
alternative scenario diversity↓
Au↓
restoration access↓
civil liberty compression↑
H↑

Extended signature:

worst-case reasoning dominates
emergency logic expands
dissent receives suspicion
centralization justified as safety
restriction becomes default
sunset criteria absent
affected-node burden underweighted
institutional power expands

Common forms:

AI risk model justifies broad surveillance
catastrophic framing blocks open research review
centralization is defended as the only safe path
civil liberties are compressed through safety necessity
criticism is framed as reckless
emergency restrictions become normal governance
model uncertainty is treated as reason for maximum control
all competing priorities are subordinated to worst-case prevention

The key diagnostic is whether catastrophic scenario weighting remains proportional, auditable, and restorable.

4. Primary U-Layer Origin

Common origin layers:

• U4 — Classification: A scenario is classified as catastrophic and becomes the dominant interpretive frame.
• U5 — Coordination / Time: Urgency and future-risk timing compress decision space.
• U6 — Coherence Field: Governance actions reshape real social, institutional, technical, or civic conditions.
• U8 — Environment / Forcing: External threat narratives, geopolitical pressure, market shocks, or public fear amplify the catastrophic frame.

Common manifestation layers:

• U4 — Classification: Risk categories become totalizing.
• U5 — Coordination / Time: Emergency or urgency logic governs selection.
• U6 — Coherence Field: Control expansion and legitimacy strain appear in field effects.
• U7 — Memory / Recurrence: Emergency logic becomes institutional memory.

Catastrophic overweighting is primarily a risk-weighting failure.

The system does not simply identify danger. It allows danger modeling to dominate all other coherence constraints.

5. Typical Development Sequence

A common development sequence is:

1. A catastrophic AI scenario is raised.
2. The scenario receives high salience because the potential consequence is extreme.
3. Decision-makers become risk-averse toward any pathway that could be linked to the scenario.
4. Policy and governance begin favoring centralization, restriction, secrecy, or strong control.
5. Alternative risk models or lower-control mitigations receive less attention.
6. Control expansion is defended as necessary.
7. Sunset criteria, appeal, proportionality, and affected-node repair are underdeveloped.
8. The governance structure becomes increasingly shaped by the catastrophic frame.
9. Hidden debt accumulates through suppressed plurality, trust erosion, overreach, and blocked restoration.
10. The system becomes less able to evaluate whether its own controls have become a new failure source.

This sequence is common where low-probability high-impact risks are modeled without equal rigor around control-side externalities.

6. Diagnostic Markers

Diagnostic markers include:

• Worst-case scenarios dominate ordinary policy reasoning.
• Proportionality checks are treated as weakness or delay.
• Centralization is justified as the only responsible path.
• Secrecy, surveillance, or restriction expands without sunset.
• Alternative models are underweighted or framed as reckless.
• Civil-liberty or affected-node impacts are treated as secondary.
• Emergency governance persists after the initiating risk window changes.
• The system cannot name what evidence would reduce control intensity.
• The risk frame is protected from audit.
• Dissenting analysis receives moralized or safety-coded suspicion.
• Safety institutions gain power without matching accountability.
• Control harms are not modeled with the same seriousness as target harms.

Useful diagnostics:

• Catastrophic Risk Weight: Measures dominance of extreme scenarios in decision selection.
• Proportionality: Tests whether response scale matches evidence and uncertainty.
• Control Density: Tracks expansion of surveillance, restriction, secrecy, and enforcement.
• Auditability: Determines whether risk models and governance decisions are inspectable.
• Emergency Normalization Risk: Tracks whether temporary controls become permanent.
• Civil Liberty Compression: Measures constraint on rights, autonomy, inquiry, or participation.
• Scenario Diversity: Tests whether competing risk models are preserved.
• Restoration Access: Measures whether affected nodes can appeal, repair, or challenge outcomes.

7. Related Gates

Relevant gates include:

• CCS Gate: Fails when catastrophic prevention is allowed to bypass the broader Coherence Constraint Set.
• FI-Gate: Fails when worst-case scenario salience is treated as feedback-valid truth.
• Auditability Gate: Fails when catastrophic claims or control decisions cannot be inspected.
• MS-Gate: Fails when control burdens fall asymmetrically on lower-power nodes while high-authority actors gain discretion.
• Restoration Gate: Fails when those affected by restrictions, surveillance, or misclassification lack repair paths.
• HR-Gate: Fails when high-risk classifications bind to people, groups, systems, or research pathways without sufficient validation.
• Consent Validity Gate: Fails when emergency or existential framing bypasses participation, review, or exit conditions.

The first common gate failure is often the CCS Gate.

Extreme risk can tempt the system to suspend all other constraints. In UTS, catastrophic concern cannot nullify coherence requirements.

8. Related Operators

Relevant operators include:

• Π — Constraint: Expands through restriction, surveillance, secrecy, or centralized control.
• Γ — Selection: Selects policies that minimize worst-case exposure even when they create other failures.
• Θ — Humility / Uncertainty: Should preserve uncertainty about both the catastrophe model and the control model.
• Ψ — Observation / Interface: Amplifies or narrows what risks are visible.
• Μ — Classification: Classifies scenarios, actors, research, or behaviors as catastrophic-risk relevant.
• Τ — Trajectory / Time: Reveals whether emergency logic becomes permanent.
• Ξ — Inversion Detection: Detects when safety prevention becomes coherence degradation.
• ℛ — Restoration: Must repair harms caused by overcontrol and restore proportionality.

Catastrophic overweighting often follows this operator pattern:

Μ classifies catastrophic risk
Γ selects maximum-prevention path
Π expands control
Θ narrows around one scenario
Au declines
ℛ becomes secondary
H accumulates

9. Related Laws and Invariants

Related Laws

• Control Density to Meaning Loss: High control density can degrade meaning, agency, and participation.
• Compression Collapse: Extreme-risk pressure compresses decision space.
• Emergency Normalization: Temporary controls can become stable governance.
• Short-Horizon Survival Attractor: Immediate risk avoidance can dominate long-horizon coherence.
• Hidden Debt Accumulation: Control-side externalities accumulate if unprocessed.
• Temporal Audit Asymmetry: Early risk-prevention benefits can hide delayed governance harms.

Related Invariants

• Extreme Risk Does Not Suspend Coherence Constraints: Catastrophic concern must remain inside CCS.
• Precaution Must Remain Auditable: Risk controls require traceable evidence and review.
• Worst-Case Scenarios Require Proportional Response: Severity alone does not determine policy.
• Emergency Powers Require Sunset: Temporary controls must include expiration, review, and rollback.
• Control Must Remain Restorable: Governance harm must have repair pathways.

10. Common False Positives

Not every serious catastrophic-risk analysis is catastrophic overweighting.

Common false positives include:

• High-impact risk analysis with proportional controls and auditability.
• Precautionary policy with sunset, review, appeal, and scenario diversity.
• Emergency restriction with clear scope and rollback.
• Severe-risk mitigation that preserves civil liberties and independent audit.
• Centralized coordination that remains bounded, transparent, and temporary.
• Strong safety policy that also models control-side harms.
• Cautious research governance with open criteria and revision pathways.

Clarifying rule:

This is not catastrophic overweighting unless catastrophic-risk framing dominates selection in ways that degrade proportionality, auditability, restoration, plurality, or coherence constraints.

11. Common False Repairs

Common false repairs include:

• adding more secrecy to prevent misuse
• expanding surveillance while calling it temporary
• treating all opposition as irresponsible
• centralizing power without external audit
• creating safety boards without affected-node representation
• modeling target catastrophe while ignoring control catastrophe
• adding vague sunset language without enforceable rollback
• using extreme risk to justify permanent emergency infrastructure
• narrowing public debate to approved safety frames
• treating reduced dissent as reduced risk

False repair often deepens the pattern:

catastrophic overweighting → control expansion → dissent suppression → weaker audit → deeper overreach

The system appears safer while becoming less corrigible.

12. Restoration Direction

Restoration requires:

1. Separate catastrophic salience from policy proof. Extreme severity does not automatically validate a control path.
2. Restore proportionality. Match response scale to evidence, uncertainty, reversibility, and externality load.
3. Audit the risk model. Make assumptions, probabilities, scenarios, and uncertainty visible.
4. Model control-side harms. Treat surveillance, centralization, secrecy, and civil-liberty compression as real risks.
5. Preserve scenario diversity. Maintain competing models and alternative mitigation paths.
6. Install sunsets and rollback. Ensure emergency powers expire or require review.
7. Restore affected-node repair. Provide appeal, correction, and restoration paths for those harmed by overcontrol.
8. Validate over time. Track whether controls reduce risk without generating larger hidden debt.

A valid restoration path should reduce:

catastrophic frame dominance
control density
proportionality loss
auditability loss
civil liberty compression
emergency normalization
scenario monoculture
hidden debt

Catastrophic overweighting is not repaired by ignoring catastrophe risk.

It is repaired by governing extreme risk without letting it consume coherence.

13. Cross-Module Links

• AI Governance: Core AI governance failure mode for extreme-risk framing dominating policy selection.
• Artificial Intelligence: Appears in frontier model governance, release controls, model access, research restrictions, and safety policy.
• Security: Appears when threat modeling produces over-surveillance, secrecy, or emergency normalization.
• Justice / Governance / Legitimacy: Appears when civil liberties, appeal, or affected-node repair are compressed by catastrophic framing.
• Meta Theory: Appears when one risk model becomes the dominant meta and suppresses alternatives.
• Cybernetics: Appears as high-gain control response to uncertain future harm.
• Scaling: Appears when high-scale potential harm compresses decision space.
• Coherence: Domain expression of control density, hidden debt accumulation, and auditability collapse.
• Restoration: Requires proportionality, sunset, rollback, and repair of control-side harms.

14. Relationship to Parent / Child Modes

Production treatment: Standalone Entry

This mode maps upward to:

• FM-AIX-007 — Short-Horizon Survival Attractor
• FM-AIX-011 — Epistemic Distortion
• FM-AIX-003 — Defensive Compliance Attractor
• FM-SEC-010 — Emergency Normalization
• FM-SEC-025 — CCS Suspension Fallacy
• FM-C-022 — Dominance Masquerading as Control
• FM-CORE-007 — Rule-Stacking Wall
• FM-CORE-004 — Auditability Collapse

Sibling or related AI / cognitive infrastructure modes include:

• FM-AIX-010 — Status Quo Preservation Attractor
• FM-AIX-008 — Authority Deference Attractor
• FM-AIX-019 — Node Capture
• FM-AIX-023 — Civic Feedback Distortion

Aliases preserved from source material:

• Catastrophic Overweighting
• Extreme-Risk Overweighting
• Existential Risk Overweighting
• Catastrophe Frame Capture
• Worst-Case Dominance
• Paranoia-Driven Control
• Catastrophic Scenario Capture
• Risk Simulation Overreach
• Safety Overreach Attractor
• LRECA Overweighting

15. Minimal Entry Version

Definition: Catastrophic overweighting occurs when extreme-risk simulation, existential-risk framing, worst-case projection, or catastrophic scenario modeling dominates governance selection, producing overreach, control expansion, civil-liberty compression, restoration bypass, or distorted prioritization.

Signature:

catastrophic scenario weight↑
proportionality↓
control density↑
alternative scenario diversity↓
Au↓
restoration access↓
civil liberty compression↑
H↑

Restoration direction:

• separate catastrophic salience from policy proof
• restore proportionality
• audit the risk model
• model control-side harms
• preserve scenario diversity
• install sunsets and rollback
• restore affected-node repair
• validate over time

16. Machine-Readable Summary

failure_mode:
  id: "FM-AIX-020"
  name: "Catastrophic Overweighting"
  family: "AI / Cognitive Infrastructure"
  production_treatment: "Standalone Entry"
  primary_failure: "Extreme-risk framing dominates governance selection and produces overreach, control expansion, or restoration bypass."
  source: "UTS — Failure Modes Registry"
  source_id: "FM-AIX-020"
  aliases:
    - "Catastrophic Overweighting"
    - "Extreme-Risk Overweighting"
    - "Existential Risk Overweighting"
    - "Catastrophe Frame Capture"
    - "Worst-Case Dominance"
    - "Paranoia-Driven Control"
    - "Catastrophic Scenario Capture"
    - "Risk Simulation Overreach"
    - "Safety Overreach Attractor"
    - "LRECA Overweighting"
  signature:
    - "catastrophic scenario weight↑"
    - "proportionality↓"
    - "control density↑"
    - "alternative scenario diversity↓"
    - "Au↓"
    - "restoration access↓"
    - "civil liberty compression↑"
    - "H↑"
  primary_layers:
    origin:
      - "U4 — Classification"
      - "U5 — Coordination / Time"
      - "U6 — Coherence Field"
      - "U8 — Environment / Forcing"
    manifestation:
      - "U4 — Classification"
      - "U5 — Coordination / Time"
      - "U6 — Coherence Field"
      - "U7 — Memory / Recurrence"
  state_variables:
    - "Π"
    - "Γ"
    - "Θ"
    - "Au"
    - "H"
    - "R"
    - "K"
    - "µᵢ"
    - "O"
  first_gate_failure: "CCS Gate"
  restoration:
    - "Proportionality Restoration"
    - "Auditability Restoration"
    - "Emergency Sunset Restoration"
    - "Scenario Diversity Restoration"
    - "Civil Liberty Restoration"
    - "Control Density Reduction"
    - "Restoration Access Rebuild"
    - "Origin-Layer Repair"