Security & Compliance

Consent model

How Perceive8 collects, records, and manages user and participant consent.

Consent model

Version: 1.0
Last Updated: 2026-03-07
Owner: Privacy & Compliance Team
Classification: Internal — Confidential
Related Docs: Data inventory · Data retention · Subprocessors · Privacy requests

1. Overview

This document defines the consent model for the Perceive8 meeting intelligence platform. It specifies what consent is required for each processing activity, how consent is captured and stored, and how users can withdraw consent. It also addresses special categories of data that require enhanced consent mechanisms.

Regulatory Framework

This consent model is designed to comply with:

GDPR (EU General Data Protection Regulation) — Arts. 6, 7, 9, 22
CCPA/CPRA (California Consumer Privacy Act / California Privacy Rights Act)
BIPA (Illinois Biometric Information Privacy Act)
PIPEDA (Canada Personal Information Protection and Electronic Documents Act)
State-level recording consent laws (varies by jurisdiction)

2. Lawful Basis for Processing

2.1 Processing Activities & Lawful Basis

Processing Activity	Lawful Basis (GDPR)	Consent Type	Required	Granular Opt-Out
Account creation & authentication	Contract performance (Art. 6(1)(b))	Terms of Service acceptance	Yes — mandatory	No
Audio file upload & storage	Consent (Art. 6(1)(a))	Explicit upload action	Yes — mandatory for feature	N/A
Transcription (OpenAI Whisper, Replicate)	Consent (Art. 6(1)(a))	Included in AI processing consent	Yes — mandatory for feature	No (core feature)
Speaker diarization (Pyannote.ai, Replicate)	Consent (Art. 6(1)(a))	Included in AI processing consent	Yes — mandatory for feature	No (core feature)
Voiceprint creation & storage	Explicit consent (Art. 9(2)(a))	Separate biometric consent	Yes — opt-in only	Yes — can disable
Emotion/prosody analysis (Hume AI)	Consent (Art. 6(1)(a))	Included in AI processing consent	Yes — opt-in recommended	Yes — can disable
Sentiment/entity analysis (AssemblyAI)	Consent (Art. 6(1)(a))	Included in AI processing consent	Yes — mandatory for feature	Partial (per-feature)
Content moderation (AssemblyAI)	Legitimate interest (Art. 6(1)(f))	No separate consent needed	N/A	No (safety feature)
AI chat / RAG queries (OpenAI)	Contract performance (Art. 6(1)(b))	Included in service terms	Yes — mandatory for feature	N/A
Vector embedding generation (OpenAI)	Contract performance (Art. 6(1)(b))	Included in service terms	Yes — mandatory for feature	N/A
Real-time streaming transcription	Consent (Art. 6(1)(a))	Explicit stream initiation	Yes — mandatory for feature	N/A
Billing & payment processing (Stripe)	Contract performance (Art. 6(1)(b))	Included in service terms	Yes — mandatory	No
Audit logging	Legitimate interest (Art. 6(1)(f))	No separate consent needed	N/A	No (security)
Marketing communications	Consent (Art. 6(1)(a))	Opt-in checkbox	No — optional	Yes — unsubscribe
Analytics & product improvement	Legitimate interest (Art. 6(1)(f))	Privacy policy notice	N/A	Yes — opt-out available

2.2 Legitimate Interest Assessments (LIA)

For processing activities based on legitimate interest, a Legitimate Interest Assessment must be documented:

Activity	Legitimate Interest	Necessity	Balancing Test	Override Risk
Content moderation	Platform safety, legal compliance	Essential for preventing harmful content	Low impact on data subjects; high benefit for safety	Low
Audit logging	Security, fraud prevention, compliance	Essential for incident response and regulatory compliance	Minimal additional data; high security benefit	Low
Analytics	Product improvement, service reliability	Necessary for maintaining service quality	Aggregated/anonymized where possible	Low (with anonymization)

3. Consent Types & Collection

3.1 Consent Taxonomy

Consent ID	Name	Category	Granularity	Mandatory
`C-TOS`	Terms of Service	Contract	Account-level	Yes
`C-PP`	Privacy Policy	Notice	Account-level	Yes
`C-AUDIO`	Audio Processing	Feature consent	Per-upload	Yes (for feature)
`C-AI`	AI Provider Processing	Data sharing	Account-level	Yes (for feature)
`C-VOICE`	Voiceprint Biometric	Special category	Per-profile	No (opt-in)
`C-EMOTION`	Emotion Analysis	Feature consent	Account-level	No (opt-in recommended)
`C-STREAM`	Real-Time Streaming	Feature consent	Per-session	Yes (for feature)
`C-MARKETING`	Marketing Communications	Marketing	Account-level	No (opt-in)
`C-THIRDPARTY`	Third-Party Recording	Recording consent	Per-upload	Conditional

3.2 How Consent Is Captured

Account Creation (`C-TOS`, `C-PP`)

┌─────────────────────────────────────────────────┐
│              Create Your Account                 │
│                                                  │
│  Email:    [________________________]            │
│  Password: [________________________]            │
│                                                  │
│  ☑ I agree to the Terms of Service               │
│  ☑ I have read and accept the Privacy Policy     │
│                                                  │
│  [        Create Account        ]                │
│                                                  │
│  By creating an account, you agree that your     │
│  uploaded audio will be processed by AI           │
│  providers as described in our Privacy Policy.   │
└─────────────────────────────────────────────────┘

Storage: Consent timestamp stored in Supabase Auth metadata
Implementation Status: ⚠️ Partial — ToS acceptance exists; no granular consent tracking

Audio Upload (`C-AUDIO`, `C-AI`)

┌─────────────────────────────────────────────────┐
│              Upload Audio File                   │
│                                                  │
│  ┌─────────────────────────────────┐             │
│  │     Drag & drop audio file      │             │
│  │     or click to browse          │             │
│  └─────────────────────────────────┘             │
│                                                  │
│  ⓘ Your audio will be processed by:             │
│    • OpenAI Whisper (transcription)              │
│    • Pyannote.ai (speaker identification)        │
│    • Replicate (diarization)                     │
│    • Hume AI (emotion analysis)                  │
│    • AssemblyAI (content intelligence)           │
│                                                  │
│  ☐ I confirm all participants in this recording  │
│    have consented to being recorded and analyzed │
│                                                  │
│  [        Upload & Analyze        ]              │
└─────────────────────────────────────────────────┘

Storage: Consent flag stored per audio_files record
Implementation Status: ❌ Not Implemented — no consent confirmation at upload time

Voiceprint Consent (`C-VOICE`)

┌─────────────────────────────────────────────────┐
│         Enable Speaker Recognition               │
│                                                  │
│  Perceive8 can create a voiceprint profile to    │
│  automatically identify speakers across your     │
│  recordings.                                     │
│                                                  │
│  ⚠️ This involves processing biometric data      │
│  (voice characteristics). This data is:          │
│                                                  │
│  • Stored securely in our database               │
│  • NOT shared with third parties after creation  │
│  • Deletable at any time from your settings      │
│  • Subject to a maximum 3-year retention period  │
│                                                  │
│  ☐ I consent to the creation and storage of      │
│    voiceprint profiles for speaker recognition   │
│                                                  │
│  ☐ I understand this involves biometric data     │
│    processing and I can withdraw consent at      │
│    any time                                      │
│                                                  │
│  [     Enable Voiceprints     ]  [ Cancel ]      │
└─────────────────────────────────────────────────┘

Storage: Dedicated consent record with timestamp, version, and withdrawal capability
Implementation Status: ❌ Not Implemented — voiceprints are created without explicit biometric consent

Emotion Analysis Opt-In (`C-EMOTION`)

┌─────────────────────────────────────────────────┐
│         Emotion Analysis Settings                │
│                                                  │
│  Perceive8 can analyze emotional tone and        │
│  prosody in your recordings using Hume AI.       │
│                                                  │
│  This analysis detects:                          │
│  • Emotional states (joy, anger, sadness, etc.)  │
│  • Vocal characteristics and tone                │
│  • Non-speech vocalizations                      │
│                                                  │
│  ⓘ Emotion data is linked to identified          │
│    speakers and stored with your analysis.       │
│                                                  │
│  Emotion Analysis: [  ON  |  OFF  ]              │
│                                                  │
│  [     Save Preferences     ]                    │
└─────────────────────────────────────────────────┘

Storage: User preference in account settings
Implementation Status: ❌ Not Implemented — emotion analysis runs by default on all uploads

4. Consent Storage Schema

4.1 Proposed `user_consents` Table

❌ Not Yet Implemented — This table needs to be created.

CREATE TABLE user_consents (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    user_id UUID NOT NULL,              -- References users.external_id
    consent_type VARCHAR(50) NOT NULL,  -- e.g., 'C-TOS', 'C-VOICE', 'C-EMOTION'
    consent_version VARCHAR(20) NOT NULL, -- Version of the consent text shown
    granted_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
    withdrawn_at TIMESTAMPTZ,           -- NULL if consent is active
    ip_address INET,                    -- IP at time of consent
    user_agent TEXT,                     -- Browser/client at time of consent
    consent_text_hash VARCHAR(64),      -- SHA-256 of the consent text shown
    metadata JSONB,                     -- Additional context (e.g., upload_id for per-upload consent)
    created_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
    updated_at TIMESTAMPTZ NOT NULL DEFAULT NOW()
);

CREATE INDEX idx_user_consents_user_id ON user_consents(user_id);
CREATE INDEX idx_user_consents_type ON user_consents(consent_type);
CREATE INDEX idx_user_consents_active ON user_consents(user_id, consent_type) WHERE withdrawn_at IS NULL;

4.2 Consent Record Requirements

Each consent record must capture:

Who — User ID
What — Consent type and version
When — Timestamp of grant/withdrawal
How — IP address, user agent, UI element
Which version — Hash of the consent text displayed
Context — Any relevant metadata (e.g., which upload triggered the consent)

5. Consent Withdrawal

5.1 Withdrawal Mechanisms

Consent Type	Withdrawal Method	Effect	Timeline
`C-TOS`	Account deletion	Full account and data deletion	30 days (GDPR)
`C-PP`	Account deletion	Full account and data deletion	30 days (GDPR)
`C-AUDIO`	Delete individual analysis	Removes specific audio and derived data	Immediate
`C-AI`	Account settings toggle	Stops future AI processing; existing data retained	Immediate (prospective)
`C-VOICE`	Account settings → Delete voiceprints	Deletes all voiceprint profiles and embeddings	Immediate
`C-EMOTION`	Account settings toggle	Disables emotion analysis for future uploads	Immediate (prospective)
`C-STREAM`	End streaming session	Stops real-time processing	Immediate
`C-MARKETING`	Unsubscribe link / account settings	Stops marketing communications	Immediate
`C-THIRDPARTY`	N/A (per-upload)	Cannot retroactively withdraw (data already processed)	N/A

5.2 Withdrawal Flow

User requests consent withdrawal
         │
         ▼
┌─────────────────────┐
│  Validate request    │
│  (auth + consent ID) │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐     ┌──────────────────────┐
│  Update consent      │────▶│  Trigger side effects │
│  record              │     │  (data deletion, etc.) │
│  (set withdrawn_at)  │     └──────────┬───────────┘
└─────────────────────┘                 │
                                        ▼
                              ┌──────────────────────┐
                              │  Send confirmation    │
                              │  email to user        │
                              └──────────────────────┘

5.3 Implementation Status

Withdrawal Feature	Status	Notes
Account deletion (full withdrawal)	❌ Not Implemented	See Deletion & Export Workflow
Per-analysis deletion	⚠️ Partial	API exists but no cascade to all derived data
Voiceprint deletion	❌ Not Implemented	No ChromaDB deletion mechanism
Emotion analysis toggle	❌ Not Implemented	No per-user feature flags
Marketing unsubscribe	❌ Not Implemented	No marketing system exists yet
Consent record storage	❌ Not Implemented	`user_consents` table not created

6. Special Categories of Data

6.1 Biometric Data (Voiceprints)

Applicable Regulations:

GDPR Art. 9 — Special categories of personal data
BIPA (Illinois) — Biometric Information Privacy Act
CCPA/CPRA — Sensitive personal information
Texas CUBI — Capture or Use of Biometric Identifier Act
Washington State Biometric Privacy Law

Requirements:

Requirement	GDPR	BIPA	Status
Explicit, informed consent before collection	✅ Required	✅ Required	❌ Not Implemented
Written policy on retention and destruction	✅ Required	✅ Required	⚠️ This document (partial)
Maximum retention period	Not specified (purpose limitation)	3 years or when purpose fulfilled	❌ Not Enforced
Right to deletion	✅ Required	✅ Required	❌ Not Implemented
No sale or profit from biometric data	N/A	✅ Required	✅ Compliant (not sold)
Reasonable security measures	✅ Required	✅ Required	⚠️ Partial (encrypted at rest)
Private right of action for violations	N/A	✅ Yes ($1,000–$5,000 per violation)	⚠️ High risk

Required Actions:

Implement explicit biometric consent flow (C-VOICE)
Create biometric data retention policy (max 3 years)
Implement voiceprint deletion from PostgreSQL and ChromaDB
Add biometric consent to user_consents table
Display BIPA-compliant notice before voiceprint creation
Allow users to opt out of voiceprint creation entirely

🔴 Critical Risk: Processing voiceprints without BIPA-compliant consent exposes the platform to statutory damages of $1,000–$5,000 per violation. This must be resolved before any Illinois users create voiceprint profiles.

6.2 Healthcare Recordings

Scenario: Users upload recordings of medical consultations, therapy sessions, or health-related meetings.

Applicable Regulations:

HIPAA (US) — Health Insurance Portability and Accountability Act
GDPR Art. 9 — Health data as special category

Risk Assessment:

Risk	Severity	Mitigation
Transcripts contain PHI (Protected Health Information)	🔴 High	Perceive8 is NOT a HIPAA-covered entity or business associate by default
Emotion analysis on therapy sessions	🔴 High	Sensitive mental health inferences
AI providers process health data	🔴 High	No BAAs (Business Associate Agreements) with AI providers

Current Position:

Perceive8 does not position itself as HIPAA-compliant
Terms of Service should prohibit uploading PHI without a BAA in place
No BAAs exist with any AI provider

Required Actions:

Add Terms of Service clause prohibiting PHI uploads without BAA
Implement content warning when healthcare-related content is detected
Evaluate BAA availability with each AI provider for future healthcare tier
Consider PII/PHI detection and redaction feature

6.3 Legal Conversations

Scenario: Users upload recordings of legal consultations, depositions, or privileged communications.

Risk Assessment:

Risk	Severity	Mitigation
Attorney-client privilege may be waived by third-party processing	🔴 High	User responsibility; warn in ToS
Legal discovery implications	🟡 Medium	Retention schedule must be defensible
Confidential legal strategy exposed to AI providers	🔴 High	No confidentiality agreements with AI providers

Required Actions:

Add Terms of Service warning about legal privilege implications
Consider "legal hold" feature to prevent auto-deletion during litigation
Document that AI provider processing may affect privilege claims

6.4 Children's Voices

Applicable Regulations:

COPPA (US) — Children's Online Privacy Protection Act (under 13)
GDPR Art. 8 — Conditions for child's consent (under 16 in most EU states)
UK Age Appropriate Design Code

Current Position:

Perceive8 Terms of Service require users to be 18+
However, uploaded recordings may contain children's voices as participants

Risk Assessment:

Risk	Severity	Mitigation
Voiceprints of minors	🔴 High	Must not create voiceprint profiles for minors
Emotion analysis of children	🟡 Medium	Ethical concerns; no specific legal prohibition
Transcription of children's speech	🟢 Low	Incidental; user responsibility

Required Actions:

Add Terms of Service clause requiring users to ensure no minors' data is processed for voiceprints
Consider age detection/warning for voiceprint creation
Document that platform is not designed for processing children's data

6.5 Employee Monitoring

Scenario: Employers upload recordings of employee meetings, performance reviews, or workplace conversations.

Applicable Regulations:

GDPR Art. 88 — Processing in the context of employment
Various state/national employment privacy laws
Works council / union consultation requirements (EU)

Risk Assessment:

Risk	Severity	Mitigation
Emotion analysis used for employee evaluation	🔴 High	Potential discrimination; EU AI Act implications
Surveillance without employee consent	🔴 High	Varies by jurisdiction
Automated decision-making about employees	🟡 Medium	GDPR Art. 22 right to human review

Required Actions:

Add enterprise agreement template with employee monitoring provisions
Implement "employer use" flag that triggers additional consent requirements
Consider restricting emotion analysis for identified employee monitoring use cases
Document EU AI Act implications for emotion recognition in workplace settings

7. Consent Flow Diagrams

7.1 New User Registration Flow

                    ┌──────────────┐
                    │  User visits  │
                    │  signup page  │
                    └──────┬───────┘
                           │
                           ▼
                    ┌──────────────┐
                    │  Display ToS  │
                    │  + Privacy    │
                    │  Policy       │
                    └──────┬───────┘
                           │
                    ┌──────┴───────┐
                    │  User accepts │──── No ───▶ Cannot create account
                    │  ToS + PP?    │
                    └──────┬───────┘
                           │ Yes
                           ▼
                    ┌──────────────┐
                    │  Store C-TOS  │
                    │  + C-PP       │
                    │  consent      │
                    └──────┬───────┘
                           │
                           ▼
                    ┌──────────────┐
                    │  Account      │
                    │  created      │
                    └──────┬───────┘
                           │
                           ▼
                    ┌──────────────┐
                    │  Optional:    │
                    │  Marketing    │
                    │  opt-in       │
                    └──────────────┘

7.2 Audio Upload Flow

                    ┌──────────────┐
                    │  User selects │
                    │  audio file   │
                    └──────┬───────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Display AI       │
                    │  processing info  │
                    │  + provider list  │
                    └──────┬───────────┘
                           │
                    ┌──────┴───────────┐
                    │  Third-party      │
                    │  recording        │──── No ───▶ Skip C-THIRDPARTY
                    │  consent needed?  │
                    └──────┬───────────┘
                           │ Yes
                           ▼
                    ┌──────────────────┐
                    │  Confirm all      │
                    │  participants     │──── No ───▶ Block upload
                    │  consented?       │
                    └──────┬───────────┘
                           │ Yes
                           ▼
                    ┌──────────────────┐
                    │  Store C-AUDIO    │
                    │  + C-AI consent   │
                    │  (if first time)  │
                    └──────┬───────────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Check voiceprint │
                    │  consent status   │
                    └──────┬───────────┘
                           │
              ┌────────────┴────────────┐
              │                         │
        C-VOICE granted           C-VOICE not granted
              │                         │
              ▼                         ▼
     ┌────────────────┐        ┌────────────────┐
     │  Process with   │        │  Process without│
     │  voiceprints    │        │  voiceprints    │
     └────────────────┘        └────────────────┘
              │                         │
              └────────────┬────────────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Check emotion    │
                    │  consent status   │
                    └──────┬───────────┘
                           │
              ┌────────────┴────────────┐
              │                         │
        C-EMOTION granted         C-EMOTION not granted
              │                         │
              ▼                         ▼
     ┌────────────────┐        ┌────────────────┐
     │  Include Hume   │        │  Skip Hume AI   │
     │  AI analysis    │        │  analysis        │
     └────────────────┘        └────────────────┘
              │                         │
              └────────────┬────────────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Begin analysis   │
                    │  pipeline         │
                    └──────────────────┘

7.3 Voiceprint Enrollment Flow

                    ┌──────────────┐
                    │  User enables │
                    │  voiceprints  │
                    └──────┬───────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Display BIPA     │
                    │  notice +         │
                    │  biometric info   │
                    └──────┬───────────┘
                           │
                    ┌──────┴───────────┐
                    │  User provides    │
                    │  explicit consent │──── No ───▶ Feature disabled
                    │  (two checkboxes) │
                    └──────┬───────────┘
                           │ Yes
                           ▼
                    ┌──────────────────┐
                    │  Store C-VOICE    │
                    │  consent with:    │
                    │  - Timestamp      │
                    │  - IP address     │
                    │  - Consent text   │
                    │    version hash   │
                    └──────┬───────────┘
                           │
                           ▼
                    ┌──────────────────┐
                    │  Enable voiceprint│
                    │  processing for   │
                    │  future uploads   │
                    └──────────────────┘

8. Recording Consent (Multi-Party)

8.1 The Recording Consent Problem

Perceive8 processes audio recordings that may contain multiple speakers. The uploader consents to processing, but other participants in the recording may not have consented.

8.2 Responsibility Model

Party	Responsibility
Perceive8 (Data Processor)	Provide tools and notices to help users comply with recording consent laws
User (Data Controller)	Ensure all recording participants have consented to recording and AI analysis
Recording Participants (Data Subjects)	Have the right to be informed and to object

8.3 Jurisdictional Requirements

Jurisdiction	Consent Standard	Requirement
US — Federal	One-party consent	Uploader's consent sufficient
US — California, Illinois, others	Two-party / all-party consent	All participants must consent
EU / UK	Informed consent	All participants must be informed of recording and AI processing
Canada	One-party (federal) / varies by province	At least one party must consent

8.4 Platform Safeguards

Current:

❌ No recording consent verification at upload
❌ No participant notification mechanism
❌ No jurisdiction detection

Proposed:

Add upload-time confirmation: "All participants consented to recording and AI analysis"
Provide shareable consent notice template for users to send to participants
Implement participant notification feature (email notice with opt-out)
Add jurisdiction selector to help users understand their obligations
Create "participant rights" page where recording subjects can request data access/deletion

9. Consent for AI Provider Data Sharing

9.1 Transparency Requirements

Users must be clearly informed that their audio data is sent to third-party AI providers for processing. This includes:

Information	Where Disclosed	Status
List of AI providers	Privacy Policy, Upload UI	❌ Not in upload UI
What data is sent to each provider	Privacy Policy	⚠️ Partial (in Privacy Policy)
Provider data locations (countries)	Privacy Policy	❌ Not disclosed
Provider retention policies	Privacy Policy	❌ Not disclosed
Provider DPA status	Internal only	❌ Not disclosed to users

9.2 Required Disclosures

Before first audio upload, users must be shown:

Which providers will process their data
What data is sent to each provider
Where the data is processed (country)
How long providers retain the data
Their rights regarding provider-held data

See Subprocessors for the full provider registry.

10. Implementation Status Summary

Consent Infrastructure

Component	Status	Priority
`user_consents` database table	❌ Not Implemented	🔴 Critical
Consent capture at registration	⚠️ Partial (ToS only)	🔴 Critical
Consent capture at upload	❌ Not Implemented	🔴 Critical
Biometric consent flow (voiceprints)	❌ Not Implemented	🔴 Critical
Emotion analysis opt-in/opt-out	❌ Not Implemented	🟡 High
Consent withdrawal mechanism	❌ Not Implemented	🔴 Critical
Consent version tracking	❌ Not Implemented	🟡 High
Marketing consent (opt-in)	❌ Not Implemented	🟢 Standard
Recording consent verification	❌ Not Implemented	🟡 High
Provider disclosure in upload UI	❌ Not Implemented	🟡 High
Participant notification system	❌ Not Implemented	🟢 Standard

Implementation Roadmap

Phase 1 — Critical (Before EU/BIPA exposure):

Create user_consents table
Implement biometric consent flow for voiceprints
Add consent withdrawal mechanism
Add upload-time consent confirmation

Phase 2 — High Priority (Enterprise readiness): 5. Implement emotion analysis opt-in toggle 6. Add provider disclosure to upload UI 7. Implement consent version tracking 8. Add recording consent verification

Phase 3 — Standard (Full compliance): 9. Build participant notification system 10. Implement marketing consent 11. Add jurisdiction-aware consent flows 12. Create consent audit dashboard

11. Revision History

Version	Date	Author	Changes
1.0	2026-03-07	Privacy & Compliance Team	Initial consent model

← Back to docs

Consent model

1. Overview

Regulatory Framework

2. Lawful Basis for Processing

2.1 Processing Activities & Lawful Basis

2.2 Legitimate Interest Assessments (LIA)

3. Consent Types & Collection

3.1 Consent Taxonomy

3.2 How Consent Is Captured

Account Creation (C-TOS, C-PP)

Audio Upload (C-AUDIO, C-AI)

Voiceprint Consent (C-VOICE)

Emotion Analysis Opt-In (C-EMOTION)

4. Consent Storage Schema

4.1 Proposed user_consents Table

4.2 Consent Record Requirements

5. Consent Withdrawal

5.1 Withdrawal Mechanisms

5.2 Withdrawal Flow

5.3 Implementation Status

6. Special Categories of Data

6.1 Biometric Data (Voiceprints)

6.2 Healthcare Recordings

6.3 Legal Conversations

6.4 Children's Voices

6.5 Employee Monitoring

7. Consent Flow Diagrams

7.1 New User Registration Flow

7.2 Audio Upload Flow

7.3 Voiceprint Enrollment Flow

8. Recording Consent (Multi-Party)

8.1 The Recording Consent Problem

8.2 Responsibility Model

8.3 Jurisdictional Requirements

8.4 Platform Safeguards

9. Consent for AI Provider Data Sharing

9.1 Transparency Requirements

9.2 Required Disclosures

10. Implementation Status Summary

Consent Infrastructure

Implementation Roadmap

11. Revision History

Account Creation (`C-TOS`, `C-PP`)

Audio Upload (`C-AUDIO`, `C-AI`)

Voiceprint Consent (`C-VOICE`)

Emotion Analysis Opt-In (`C-EMOTION`)

4.1 Proposed `user_consents` Table