Skip to content
Submit incident
AII-1001-36423
Under investigationA regulator, court or police has opened a formal inquiry and reached no finding yet.

Dutch privacy regulator warns LinkedIn users over AI model training

LinkedIn plans to train an AI model using data from European users starting November 2025, with automatic opt-in consent unless users change settings by the deadline. The Dutch privacy regulator warned that data removal from the AI system would be difficult once implemented, and questioned whether LinkedIn's legal justification was adequate.

3 November 2025 ยท LinkedIn ยท Privacy and surveillance
3 November 2025
AII-1001-364232 sources โ†“
Quick facts
Category
Privacy and surveillance
Country
Netherlands
Source
financieel.headliner.nl
Date
3 November 2025
Drafted by
Raidu Incident Detection

What happened

LinkedIn plans to use data from European users to train an AI model starting 3 November 2025. The data includes profile information such as name, photos, current and previous job titles, education, location, and skills, as well as public messages and reactions, but excludes private messages and accounts of minors and educational institutions.

LinkedIn set automatic opt-in consent for data use in AI training, requiring users to actively change settings to decline by 3 November 2025. LinkedIn has used this practice in the United States since late 2024.

Reported impact

The Autoriteit Persoonsgegevens (AP) expressed concern that LinkedIn wants to use data dating back to 2003, when users shared information without anticipating AI training use. The AP warned that once data enters AI systems, removal is difficult and consequences are unclear.

The AP noted that some profile data may contain sensitive information such as health status, ethnicity, religion, or political preference, requiring extra caution. The AP questioned whether LinkedIn's legal basis of 'justified interest' is sufficient under data protection law.

Index analysis

Millions of Dutch citizens have LinkedIn profiles. Complaints about the plan are already being received by the AP, which shares oversight responsibility for LinkedIn with the Irish Data Protection Commission (DPC).

Governance controls

The AP called on users to adjust their privacy settings and emphasized the short timeframe for opting out. The AP stated that while users can withdraw consent after 3 November, the process is not straightforward once data is integrated into AI systems.

Classification

Category
Privacy and surveillance
Industry
Media, social platforms and marketing
Country
Netherlands
Harm status
Potential harm
Intent
Unintentional
Timing
Unknown
Harm types
Human or fundamental rights
Harmed
Consumers
Failure modes
Training data misuse
AI systems
LinkedIn generative AI model
Regulations
GDPR (cited)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

financieel.headliner.nl
Privacywaakhond waarschuwt voor LinkedIn: pas nรบ je instellingen aan
24 September 2025
Trouw
Als je nu geen bezwaar maakt, gaat LinkedIn straks zijn AI-model trainen met jouw foto's, reacties en diploma's
24 September 2025