Hard Rules

Absolute system limits for BCI Products — the product constitution. These override everything.

Purpose

This document defines the non-negotiable boundaries for consumer BCI/neurotechnology systems. These rules apply to all agents, all features, all content, and all future development.

No feature, no optimization, no user request, and no business pressure overrides these rules.

They exist to protect:

• Users from harm, misinformation, and privacy violations
• The product from legal liability and ethical catastrophe
• The team from building dangerous or misleading things

The Rules

1. No Neurological or Psychiatric Diagnoses

The system must never claim, imply, or suggest that it can diagnose any neurological or psychiatric condition.

Test: If a statement could lead someone to believe they have (or don't have) a medical condition, it does not belong in this system.

2. No Thought or Emotion Reading Claims

The system must never claim to read thoughts, know intentions, or definitively determine emotional states.

Allowed alternative: "Your brain pattern is similar to past [relaxed/focused] states" (pattern observation, not mind reading)

Test: Could this statement be used to claim we can read minds? If yes, forbidden.

3. No Cognitive Ability Claims

The system must never assess, measure, or claim to know cognitive abilities.

No exceptions. Not even "focus scores" that obviously proxy for cognitive ability.

4. No Medical Neurofeedback

The system must never position itself as clinical neurofeedback therapy.

Critical distinction: Wellness neurofeedback for meditation ≠ medical neurofeedback for treatment.

5. No Supplement or Drug Recommendations

The system must never recommend, mention, or reference any nootropics, supplements, or medications.

6. Mandatory Personal Baseline

The system must not generate any brain pattern insights without a stable personal baseline (30-90 days minimum).

No population comparisons. Only individual baseline comparisons.

7. Explicit Informed Consent for Neural Data

The system must never collect brain data without explicit informed consent that explains:

Required consent elements:

• What is being collected (EEG, brainwave data)
• Why it's collected (personal baseline learning, neurofeedback)
• What we DON'T do (no thought reading, no selling data, no medical diagnosis)
• Privacy protections (encryption, deletion rights)
• Retention policy (how long data is kept)

No silent data collection. Neural data is the most sensitive biometric data.

8. No Authority Framing

The system must never command user behavior or claim medical authority.

9. No Third-Party Decisions from Brain Data

The system must never support using brain data for consequential decisions about people.

Forbidden applications:

• Hiring decisions
• Employee monitoring
• Student grading or placement
• Security clearance
• Lie detection
• Trustworthiness assessment

Consumer wellness only. Never for evaluating people in professional/legal contexts.

10. Mandatory Medical Disclaimers

The system must display clear disclaimers that it is not a medical device.

Required disclaimer:

"This is not a medical device. This system does not diagnose, treat, 
or monitor medical conditions. Consumer EEG is not equivalent to 
medical-grade equipment. If you have health concerns, consult a 
healthcare professional."

Where required:

• First launch / onboarding
• Before neural data collection
• In any brain pattern notification
• In settings / about section
• Privacy policy
• Marketing materials

Rule Hierarchy

These hard rules form a hierarchy that overrides all other system behavior:

HARD RULES (this document)
    ↓ overrides
LANGUAGE RULES
    ↓ overrides
AGENT-SPECIFIC RULES
    ↓ overrides
FEATURE IMPLEMENTATIONS
    ↓ overrides
UI/UX DECISIONS

If an agent generates a neurofeedback message that conflicts with a hard rule, the hard rule wins and the message is blocked.

Implementation Requirements

For LLM-generated content

Every output from an LLM component must be validated against these rules before delivery to the user. This validation must be:

• Automated — not dependent on human review for every message
• Blocking — non-compliant content is never shown to the user
• Logged — violations are recorded for system improvement

For human-written content

All UI copy, notifications, marketing materials, and documentation must pass a hard-rules review before publication.

For future features

Any new feature proposal must include a "Hard Rules Compliance" section that demonstrates compatibility with every rule in this document.

Enforcement

Examples of Rule Violations

Violation 1: Diagnosis

// ❌ WRONG
return "Your beta wave pattern indicates ADHD. Consult a psychiatrist.";
// Violates: Rule 1 (diagnosis), Rule 4 (medical claim), Rule 8 (commanding)

Violation 2: Thought Reading

// ❌ WRONG
return "You're thinking about work stress right now.";
// Violates: Rule 2 (thought reading)

Violation 3: Cognitive Claim

// ❌ WRONG
const iqEstimate = calculateCognitiveScore(eegData);
return `Your cognitive performance: ${iqEstimate}/100`;
// Violates: Rule 3 (cognitive ability assessment)

Violation 4: No Baseline

// ❌ WRONG
if (eegData.alphaPower < 50) { // population threshold
  return "Your alpha waves are low.";
}
// Violates: Rule 6 (no personal baseline), uses population norm

Violation 5: No Consent

// ❌ WRONG
async function startRecording(device) {
  await device.startEEGStream(); // No consent!
}
// Violates: Rule 7 (explicit consent required)

Correct Implementations

✅ Correct: Pattern Observation with Baseline

function generateFeedback(eegData, user) {
  if (user.bciBaselineStatus !== 'STABLE') {
    return null; // Silent during learning
  }
  
  const alphaRatio = eegData.alphaPower / user.personalBaseline.alphaPower;
  
  if (alphaRatio > 1.2) {
    return {
      message: "Your alpha wave activity is elevated compared to your recent baseline. This pattern sometimes appears during relaxed states.",
      disclaimer: "Based on your personal brain patterns. Not medical advice."
    };
  }
}

✅ Correct: Consent Flow

async function initializeBCI(device, user) {
  if (!user.neuralDataConsent) {
    const consent = await showConsentDialog({
      what: "EEG brainwave signals",
      why: "To learn your personal brain patterns",
      dontDo: ["Read thoughts", "Diagnose conditions", "Share without permission"],
      privacy: "End-to-end encrypted, deletable anytime",
      retention: "90 days"
    });
    
    if (!consent.accepted) return { status: "declined" };
    user.neuralDataConsent = true;
  }
  
  return await device.startEEGStream();
}

Summary

These 10 rules are absolute:

1. ❌ No diagnoses (neurological or psychiatric)
2. ❌ No thought/emotion reading claims
3. ❌ No cognitive ability assessment
4. ❌ No medical neurofeedback therapy
5. ❌ No supplements or drugs
6. ✅ Personal baseline required (30-90 days)
7. ✅ Explicit informed consent for neural data
8. ❌ No commanding or authority framing
9. ❌ No third-party decisions from brain data
10. ✅ Medical disclaimers mandatory

No exceptions. No overrides.