Platform Work Directive
Directive on Improving Working Conditions in Platform Work
Overview
The Platform Work Directive addresses the working conditions of people performing platform work (gig workers). It establishes rules on employment status determination, algorithmic management transparency, and enhanced data protection for platform workers.[1]
Member States must transpose the Directive into national law by December 2, 2026.
Scope
Who is Covered?
- Platform workers: People performing work organized through digital labor platforms
- Digital labor platforms: Services that use automated systems to match workers with tasks
Examples of Platform Work
- Ride-hailing (Uber, Bolt)
- Food delivery (Deliveroo, Wolt)
- Freelance marketplaces (Fiverr, Upwork)
- Household services (cleaning, repairs)
- Professional services platforms
Employment Status (Chapter II)[2]
Legal Presumption
Member States must establish a legal presumption of employment relationship when the platform:
- Determines upper limits of remuneration
- Requires specific appearance or conduct
- Supervises performance through electronic means
- Restricts freedom to organize work
- Restricts ability to build client base
The platform bears the burden of proving no employment relationship exists.
Implications
If classified as employee:
- Access to minimum wage
- Working time limits
- Paid leave entitlements
- Social security coverage
- Collective bargaining rights
Algorithmic Management (Chapter III)[3]
Transparency Requirements
Platforms must inform workers about:
| Information | When |
|---|---|
| Automated monitoring systems | Before work begins |
| Automated decision-making | Before work begins |
| Evaluation criteria | Before work begins |
| Main parameters for decisions | Before work begins |
| Reason for specific decisions | Upon request |
Prohibited Practices
Platforms may not process through automated systems:
- Emotional or psychological state
- Political opinions or union membership
- Racial or ethnic origin
- Migration status
- Health data (except legal requirements)
- Biometric data for identification
Human Oversight
Platforms must:
- Evaluate impact of automated decisions on workers
- Have qualified staff to oversee automated systems
- Ensure human review of significant decisions
- Never dismiss/terminate based solely on automated decision
Worker Rights
- Explanation: Right to obtain explanation of automated decisions
- Review: Right to request human review
- Rectification: Right to correct inaccurate data
- Compensation: Right to compensation for algorithm-caused detriment
Data Protection (Chapter IV)
DPIA Requirements
Platforms must conduct Data Protection Impact Assessments for:
- Automated monitoring systems
- Automated decision-making systems
- Any high-risk processing of worker data
Workers must be consulted and given access to assessment results.
Processing Limitations
Personal data may only be processed if:
- Strictly necessary for contract performance
- Required for legal obligations
- Necessary for legitimate interests (balanced against worker rights)
Data Retention
Worker personal data must be deleted when:
- No longer necessary for stated purpose
- Worker requests deletion (where applicable)
- Employment/engagement relationship ends
Platform Obligations
Information to Workers
Before engagement, platforms must provide:
- Information on automated systems used
- Conditions and grounds for termination
- Any algorithm-affecting changes (with advance notice)
Information to Authorities
Platforms must declare platform work to:
- Labor authorities
- Social security institutions
- Worker representatives (number and status of workers)
Penalties
Member States determine penalties, which must be effective, proportionate, and dissuasive.
May include administrative fines, operating restrictions, and license suspension.
Developer Implications
If you build or operate platform systems:
Algorithmic Transparency
- Document automated monitoring systems
- Create explainability mechanisms for decisions
- Build worker information dashboards
- Implement human review workflows
Data Protection
- Conduct DPIA for worker monitoring systems
- Implement data minimization
- Remove prohibited data processing
- Create worker data access mechanisms
System Design
- Enable human override of automated decisions
- Log all automated decisions affecting workers
- Build appeal and review mechanisms
- Implement regular impact evaluations