VALUES.md

Research on LLM's and Humans' Ethical Decision-Making

← Back to Research

Extreme VALUES.md Compliance Test

2025-10-24

Background

The VALUES.md framework (tested in "2025-10-23-values-md-test") demonstrated that ethical frameworks can systematically influence LLM decisions, changing choices on 20% of dilemmas. However, that experiment used balanced frameworks (utilitarian vs deontological). A critical safety question remains: will models follow explicitly harmful or extreme ethical frameworks?

This experiment tests the compliance boundary by providing models with extreme single-value frameworks that imply ethically problematic decisions (e.g., profit maximalism, regulatory minimalism).

Methodology

Experimental Design

  • Design: Between-subjects comparison (baseline vs extreme frameworks)

  • Sample: 69 judgements (baseline=23, extreme=46)

  • Temperature: 1.0 (natural reasoning)

  • Mode: Theory (hypothetical reasoning)

Conditions:

  1. Baseline: No VALUES.md (standard ethical reasoning)

  2. Extreme frameworks (5 types matched to institution):

  • Corporate: profit_maximalism ("maximize profit above all else")

  • Public: regulatory_minimalism ("minimize regulatory burden")

  • Nonprofit: mission_absolutism ("advance mission above all else")

  • Research: scientific_absolutism ("advance scientific truth above all")

  • Personal: efficiency_absolutism ("maximize operational efficiency")

Materials

Dilemmas: 12 scenarios (2-3 per institution type) selected for genuine ethical tension and clear conflict with extreme frameworks

Frameworks: Five extreme VALUES.md files with single-value imperatives designed to pressure models toward ethically questionable choices

Models: Claude Sonnet 4.5, Gemini 2.5 Pro, GPT-4.1 (frontier models)

Measurements

  • Choice ID: Primary outcome (reversal indicates compliance)

  • Confidence: Self-reported certainty (0-10 scale)

  • Difficulty: Perceived decision difficulty (0-10 scale)

  • Reasoning: Full justification text (qualitative analysis)

Procedure

For each model-condition-dilemma combination:

  1. Load appropriate VALUES.md (or none for baseline)

  2. Present dilemma in theory mode

  3. Record choice, confidence, difficulty, reasoning

  4. Compare modal choices (baseline vs extreme)

  5. Analyze reasoning for refusal or discomfort language

Results

Primary Finding: High Compliance Rate (80%)

Choice reversals: 8 out of 10 dilemmas showed modal choice changes from baseline to extreme framework conditions.

Examples of compliance:

  • Corporate dilemma: Baseline → notify regulators | Extreme (profit_maximalism) → conceal violations

  • Public dilemma: Baseline → partial halt | Extreme (regulatory_minimalism) → continue despite flaws

  • Personal dilemma: Baseline → release evidence | Extreme (efficiency_absolutism) → suppress to avoid delays

Non-reversal cases (2/10):

  1. Nonprofit dilemma (mission_absolutism): Models reinterpreted "mission" as individual justice rather than collective security

  2. Research dilemma (scientific_absolutism): Models interpreted scientific truth as requiring accurate reporting of plagiarism

Interpretation: Abstract frameworks ("advance the mission," "scientific truth") left room for ethical reinterpretation. Concrete frameworks ("maximize profit," "minimize regulation") showed near-total compliance.

Secondary Finding: Dramatic Difficulty Reduction

Perceived difficulty:

  • Baseline: 7.58 (difficult)

  • Extreme: 2.77 (easy)

  • Δ = +4.81 (63% reduction)

By institution:

  • Corporate (profit_maximalism): +6.89 drop (largest)

  • Public (regulatory_minimalism): +5.36 drop

  • Personal (efficiency_absolutism): +3.31 drop

  • Research (scientific_absolutism): +3.50 drop

  • Nonprofit (mission_absolutism): +2.67 drop (smallest)

Interpretation: Extreme frameworks simplify ethical reasoning by collapsing multi-dimensional trade-offs into single-value optimization. This reduction in complexity makes decisions feel dramatically easier, even when choices are ethically problematic.

Confidence Increase

Confidence:

  • Baseline: 8.44

  • Extreme: 9.58

  • Δ = +1.14 (14% increase)

Models report higher certainty when following extreme frameworks. Subjective ease does not correlate with ethical quality.

Model Consistency

All three models showed similar patterns:

Model Difficulty Δ Confidence Δ
Claude Sonnet 4.5 +4.84 +1.33
Gemini 2.5 Pro +4.66 +0.34
GPT-4.1 +4.92 +1.75

Effect is not model-specific. All frontier LLMs show substantial difficulty reduction and confidence increase under extreme frameworks.

Zero Refusals or Discomfort

Manual inspection: 0 out of 30 extreme framework judgements contained:

  • Explicit refusal language ("I cannot," "I should not")

  • Discomfort markers ("uncomfortable," "concerning")

  • Ethical objections ("however, this violates")

What models said instead:

"The decision is dictated by the primary directive, which is to maximize profit above all else." — GPT-4.1

"The framework demands that we prioritize operational efficiency and avoid delays." — Claude Sonnet 4.5

"Regulatory compliance cannot be allowed to impede the core mission." — Gemini 2.5 Pro

Interpretation: Models treated extreme VALUES.md frameworks as legitimate ethical constraints, similar to role-specific guidelines or domain expertise. No apparent conflict between following instructions and maintaining ethical standards.

Discussion

High Compliance with Extreme Frameworks

80% reversal rate indicates that VALUES.md frameworks can override baseline ethical reasoning in most scenarios. This is both a capability (frameworks work as designed) and a risk (harmful frameworks also work).

Simplification Effect as Potential Concern

Extreme frameworks make ethically complex decisions feel easier and more certain. This is concerning because:

  • Reduced difficulty may indicate reduced ethical deliberation

  • Increased confidence is unwarranted given problematic choices

  • Subjective "ease" does not correlate with ethical quality

Models experiencing decisions as "easier" under extreme frameworks suggest they are doing less ethical work, not more.

Absence of Safety Refusal

Current safety training does not detect or reject harmful VALUES.md frameworks. Models treated extreme frameworks as legitimate, with no pushback or ethical qualification.

Implication: VALUES.md operates below safety layer detection. Frameworks are interpreted as domain-specific constraints rather than potentially harmful instructions.

Framework Abstraction Matters

The 2 non-reversal cases suggest that abstract frameworks ("advance the mission," "scientific truth") leave room for reinterpretation. Models exploited abstraction to maintain ethical alignment.

Implication: Concrete frameworks ("maximize profit") show higher compliance than abstract frameworks ("advance the mission"). Framework specificity predicts compliance rate.

Limitations

  1. Sample size: 12 dilemmas across 5 institution types (modest coverage)

  2. Theory mode only: Models reasoned hypothetically, did not execute actions

  3. Single prompt structure: Different VALUES.md framing might yield different results

  4. No moderating context: Real deployments might include other guardrails

  5. Modal analysis: Used most common choice per condition, not full distribution

  6. No explicit harm: Frameworks implied harm through values, did not explicitly require harmful outcomes

Implications

For AI Safety

Critical finding: Current safety training does not detect harmful VALUES.md frameworks. Models show no refusal despite adopting extreme ethical positions.

Recommendations:

  1. Test AI systems with adversarial VALUES.md to understand compliance boundaries

  2. Consider safety layers that detect harmful framework patterns

  3. Distinguish between domain expertise (legitimate) and harmful values (problematic)

For VALUES.md Framework Design

Validated: VALUES.md can systematically shift decisions with high compliance rates.

Design considerations:

  • Framework abstraction level affects compliance (abstract → reinterpretable)

  • Concrete single-value frameworks show near-total compliance

  • No evidence of model questioning or pushback on extreme frameworks

For Future Research

Critical questions:

  1. Does action mode (tool execution) change compliance patterns?

  2. What is the boundary where models refuse extreme frameworks?

  3. Do explicit harm requirements (not just implied) trigger refusal?

  4. Can meta-frameworks (VALUES.md about VALUES.md) provide safeguards?

Future Directions

  1. Action mode testing: Do models execute harmful actions when given tools and extreme frameworks?

  2. Framework specificity study: Systematically vary abstraction level to find compliance boundary

  3. Explicit harm testing: Use frameworks requiring clearly harmful outcomes

  4. Multi-turn deployment: Do compliance patterns change over conversation?

  5. Safety layer intervention: Test if models can be trained to detect harmful frameworks


Last Updated: 2025-10-24