Installs into .claude/skills of the current project.
Are you the author of Pii Privacy Regulations?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/fabioc-aloha-pii-privacy-regulations)
---
name: pii-privacy-regulations
description: "Handling personally identifiable information under European and Australian privacy regulations."
lastReviewed: 2026-04-30
---
# PII & Privacy Regulations Skill
> Handling personally identifiable information under European and Australian privacy regulations.
## β οΈ Staleness Warning
Privacy regulations are actively evolving. Major changes expected.
**Refresh triggers:**
- GDPR enforcement updates or new guidelines
- Australian Privacy Act reform (ongoing review)
- New adequacy decisions for data transfers
- Significant enforcement actions or fines
- EU AI Act privacy provisions (2025-2026)
**Last validated:** February 2026
**Check current state:**
- EU: [GDPR.eu](https://gdpr.eu/), [EU Commission](https://commission.europa.eu/law/law-topic/data-protection_en)
- Australia: [OAIC](https://www.oaic.gov.au/privacy/australian-privacy-principles)
---
## What is PII?
Personally Identifiable Information (PII) is any data that can identify an individual directly or indirectly.
### Direct Identifiers
| Category | Examples |
| -------- | -------- |
| **Name** | Full name, maiden name, alias |
| **Government IDs** | SSN, passport, driver's license, TFN (AU) |
| **Financial** | Bank account, credit card numbers |
| **Contact** | Email, phone, physical address |
| **Biometric** | Fingerprints, facial recognition, voice |
### Indirect Identifiers (Quasi-identifiers)
| Category | Examples |
| -------- | -------- |
| **Location** | IP address, GPS coordinates, ZIP/postcode |
| **Device** | Device ID, MAC address, browser fingerprint |
| **Demographic** | Age, gender, ethnicity, occupation |
| **Behavioral** | Purchase history, browsing patterns |
> **Key insight:** Combining quasi-identifiers can uniquely identify individuals even without direct identifiers.
---
## πͺπΊ GDPR (European Union)
**Applies to:** Any organization processing EU residents' data, regardless of location.
### Core Principles (Article 5)
| Principle | Requirement |
| --------- | ----------- |
| **Lawfulness, Fairness, Transparency** | Process data legally with clear communication |
| **Purpose Limitation** | Collect only for specified, legitimate purposes |
| **Data Minimization** | Collect only what's necessary |
| **Accuracy** | Keep data accurate and up to date |
| **Storage Limitation** | Retain only as long as necessary |
| **Integrity & Confidentiality** | Ensure appropriate security |
| **Accountability** | Demonstrate compliance |
### Lawful Bases for Processing
| Basis | When to Use |
| ----- | ----------- |
| **Consent** | Freely given, specific, informed, unambiguous |
| **Contract** | Necessary for contract performance |
| **Legal Obligation** | Required by law |
| **Vital Interests** | Protect someone's life |
| **Public Task** | Official authority or public interest |
| **Legitimate Interests** | Business need balanced against rights |
### Individual Rights (Data Subject Rights)
| Right | Description | Response Time |
| ----- | ----------- | ------------- |
| **Access** | Obtain copy of their data | 1 month |
| **Rectification** | Correct inaccurate data | 1 month |
| **Erasure** ("Right to be Forgotten") | Delete their data | 1 month |
| **Restrict Processing** | Limit how data is used | 1 month |
| **Data Portability** | Receive data in machine-readable format | 1 month |
| **Object** | Stop certain processing | Without delay |
| **Automated Decision-Making** | Human review of automated decisions | 1 month |
### GDPR Compliance Checklist
```markdown
## Lawful Basis & Transparency
- [ ] Document lawful basis for each processing activity
- [ ] Maintain records of processing activities (Article 30)
- [ ] Privacy policy is clear, accessible, updated
## Data Security
- [ ] Encryption at rest and in transit
- [ ] Pseudonymization where possible
- [ ] Data Protection Impact Assessment (DPIA) for high-risk processing
- [ ] Breach notification process (72 hours to authority)
## Governance
- [ ] Data Protection Officer (DPO) appointed if required
- [ ] Data Processing Agreements with third parties
- [ ] EU representative appointed (if outside EU)
## Individual Rights
- [ ] Process to handle access requests
- [ ] Process to handle deletion requests
- [ ] Process to handle data portability requests
- [ ] Consent mechanism (opt-in, not opt-out)
```
### 2025-2026 Updates
| Update | Impact |
| ------ | ------ |
| **Simplified Record-Keeping** (May 2025) | Organizations <750 employees only need records for high-risk processing |
| **Enhanced Enforcement Procedures** (June 2025) | Additional procedural rules for cross-border cases |
| **AI Act Integration** | Additional requirements for AI systems processing personal data |
---
## π¦πΊ Australian Privacy Principles (APPs)
**Applies to:** Organizations with annual turnover >$3M AUD, government agencies, health service providers, and some others.
### The 13 APPs
| APP | Title | Key Requirement |
| --- | ----- | --------------- |
| **APP 1** | Open & Transparent Management | Have a clear, up-to-date privacy policy |
| **APP 2** | Anonymity & Pseudonymity | Allow anonymous/pseudonymous dealings where practical |
| **APP 3** | Collection of Solicited Information | Only collect necessary info; higher bar for sensitive info |
| **APP 4** | Unsolicited Personal Information | Assess and destroy if wouldn't have been collected |
| **APP 5** | Notification of Collection | Tell individuals what you're collecting and why |
| **APP 6** | Use or Disclosure | Only for primary purpose or permitted secondary purposes |
| **APP 7** | Direct Marketing | Opt-out required; sensitive info needs consent |
| **APP 8** | Cross-Border Disclosure | Ensure overseas recipients comply with APPs |
| **APP 9** | Government Identifiers | Don't adopt government IDs as your own identifier |
| **APP 10** | Quality of Information | Keep data accurate, complete, up to date |
| **APP 11** | Security of Information | Protect from misuse, loss, unauthorized access |
| **APP 12** | Access to Information | Provide access when requested |
| **APP 13** | Correction of Information | Correct inaccurate information |
### Sensitive Information (Higher Protection)
Under Australian law, sensitive information includes:
- Health information
- Genetic information
- Biometric data
- Racial or ethnic origin
- Political opinions
- Religious beliefs
- Sexual orientation
- Criminal record
- Trade union membership
> **Rule:** Sensitive information generally requires **consent** to collect.
### Notifiable Data Breaches (NDB) Scheme
**When to notify:**
- Unauthorized access, disclosure, or loss of personal information
- Likely to result in serious harm to individuals
- Remedial action hasn't prevented serious harm
**Timeline:**
- Notify OAIC and affected individuals **as soon as practicable**
- Assessment must be completed within **30 days**
### Australian Privacy Act Reform (Ongoing)
The Privacy Act is under review. Expected changes:
| Proposed Change | Status |
| --------------- | ------ |
| Expanded definition of personal information | Under review |
| New individual rights (erasure, explanation of AI decisions) | Proposed |
| Higher penalties | Enacted (up to $50M AUD) |
| Direct right of action for individuals | Under review |
| Removal of small business exemption | Under consideration |
---
## Cross-Jurisdictional Comparison
| Aspect | GDPR (EU) | APP (Australia) |
| ------ | --------- | --------------- |
| **Scope** | All organizations processing EU data | >$3M turnover + exceptions |
| **Consent** | Must be explicit opt-in | Can be implied in some cases |
| **Right to Erasure** | Explicit right | Not explicit (under review) |
| **Breach Notification** | 72 hours to authority | "As soon as practicable" |
| **Penalties** | Up to β¬20M or 4% global revenue | Up to $50M AUD |
| **Cross-Border Transfer** | Adequacy decisions, SCCs, BCRs | Must ensure APP compliance |
| **DPO Required** | Yes, in certain cases | No requirement |
---
## Code Implementation Patterns
### Never Log PII
```typescript
// β BAD
logger.info(`User ${user.email} logged in from ${user.ipAddress}`);
// β GOOD
logger.info(`User ${hashUserId(user.id)} logged in`);
```
### Encrypt PII at Rest
```typescript
// Encrypt before storing
const encryptedEmail = encrypt(user.email, encryptionKey);
await db.users.update({ id: user.id, email: encryptedEmail });
// Decrypt only when needed
const email = decrypt(storedUser.email, encryptionKey);
```
### Implement Data Minimization
```typescript
// β BAD - Fetching everything
const user = await db.users.findById(id);
return user; // Contains PII you don't need
// β GOOD - Select only needed fields
const user = await db.users.findById(id, {
select: ['id', 'displayName', 'preferences']
});
return user;
```
### Consent Management
```typescript
interface ConsentRecord {
userId: string;
purpose: 'marketing' | 'analytics' | 'personalization';
granted: boolean;
timestamp: Date;
source: 'web' | 'mobile' | 'api';
version: string; // Privacy policy version
}
// Always check consent before processing
async function canProcess(userId: string, purpose: string): Promise<boolean> {
const consent = await getLatestConsent(userId, purpose);
return consent?.granted === true;
}
```
### Data Subject Request Handler
```typescript
interface DataSubjectRequest {
type: 'access' | 'rectification' | 'erasure' | 'portability' | 'objection';
userId: string;
requestedAt: Date;
deadline: Date; // 30 days for GDPR
status: 'pending' | 'processing' | 'completed' | 'denied';
}
// Implement audit trail for all requests
async function handleDSR(request: DataSubjectRequest) {
await auditLog.record({
action: 'dsr_received',
requestType: request.type,
userId: request.userId,
timestamp: new Date()
});
// Process based on type...
}
```
### Pseudonymization Pattern
```typescript
// Replace direct identifiers with tokens
function pseudonymize(record: UserRecord): PseudonymizedRecord {
return {
pseudoId: generateToken(record.id), // Reversible with key
ageGroup: getAgeGroup(record.birthDate), // 18-25, 26-35, etc.
region: record.country, // Keep general location
// Omit: name, email, exact address, etc.
};
}
```
---
## Compliance Checklist
### For New Projects
```markdown
## Privacy Impact Assessment
- [ ] What PII will be collected?
- [ ] What is the lawful basis (GDPR) / primary purpose (APP)?
- [ ] Is all collected data necessary? (data minimization)
- [ ] How long will data be retained?
- [ ] Who will have access?
- [ ] Will data cross borders?
- [ ] What security measures are in place?
## Technical Implementation
- [ ] PII encrypted at rest
- [ ] PII encrypted in transit (TLS 1.2+)
- [ ] Logging excludes PII
- [ ] Consent captured before processing
- [ ] Data subject request endpoints implemented
- [ ] Retention/deletion automation in place
- [ ] Audit logging for PII access
```
---
## Quick Reference: When Processing is Prohibited
| Scenario | GDPR | APP |
| -------- | ---- | --- |
| No lawful basis identified | β | β |
| Sensitive data without explicit consent | β | β |
| Marketing without opt-out option | β | β |
| Cross-border transfer without safeguards | β | β |
| Retention beyond stated period | β | β |
| Collection beyond stated purpose | β | β |
---
## Resources
### Official
- [GDPR Full Text](https://gdpr.eu/tag/gdpr/)
- [EU Commission Data Protection](https://commission.europa.eu/law/law-topic/data-protection_en)
- [OAIC Australian Privacy Principles](https://www.oaic.gov.au/privacy/australian-privacy-principles)
- [OAIC Privacy Act](https://www.oaic.gov.au/privacy/privacy-legislation/the-privacy-act)
### Tools
- Data Protection Impact Assessment templates
- Consent management platforms
- Data discovery and classification tools
---