LinkedIn to train AI on public data by default with opt-out

LinkedIn will begin using some members' profiles, posts, resumes and other public activity by default to train its generative AI models starting November 3, 2025, the company said, with an opt-out available for users who do not want their future data included. The change affects multiple markets, with LinkedIn explicitly citing the European Union, the European Economic Area, the United Kingdom and Switzerland; other reporting has indicated Canada and Hong Kong may also be covered. LinkedIn said private messages will not be used and that accounts it believes belong to people under 18 will be excluded from training sets.

LinkedIn said the updates are intended to improve AI-powered features across the platform, from recruiter tools and job discovery to AI-assisted content creation and safety systems that detect harmful material. The company also revised its User Agreement and privacy notices to clarify how it will handle deepfakes, impersonation and appeals of account decisions, and to explain situations in which changes may take effect without advance notice. For European members, LinkedIn said it will rely on legitimate interests as the legal basis for processing data to train...