Overview

• Starting November 3, 2025, LinkedIn will use member data by default to improve its generative AI models unless users opt out in Settings under Data Privacy.
• For Canada, Hong Kong, the EEA, the UK and Switzerland, LinkedIn will train AI on profiles and other public content but excludes private messages from the training set.
• The AI training change does not apply in the United States and many other markets, but LinkedIn will share additional data with Microsoft and affiliates to serve personalised ads where applicable.
• Users can disable affiliate data sharing for targeted ads via the Advertising Data option in Settings, separate from the AI model improvement control.
• LinkedIn also revised its User Agreement to address deepfakes and impersonation, clarify use of secondary payment methods, outline limited notice for some changes, and detail new appeal options.