LinkedIn trained AI models on user data without prior consent
LinkedIn used personal data from its users to train generative AI models before updating its service terms to inform them of this practice, according to reports by 404 Media. The company later updated its privacy policy and terms of service to disclose the data use, while exempting users in the EU, EEA, Switzerland and the UK from this training.
What happened
LinkedIn used personal data of its users to train generative AI models before updating its service terms and informing users of this practice, according to 404 Media.
The company confirmed on 18 September 2024 that it had been automatically including user accounts in the training of generative AI models without prior notification.
LinkedIn updated its privacy policy and terms of service to disclose how it uses user information for developing AI models, with the updated terms set to take effect on 20 November 2024.
Reported impact
The data processing did not affect LinkedIn customers in the European Union, European Economic Area, Switzerland and the United Kingdom, as the company stated it does not train generative AI models with data from users in these regions.
Index analysis
LinkedIn processes personal data including user posts, usage information, language preferences and comments to train its generative AI models.
The company stated it attempts to minimize personal data in training datasets through privacy-enhancing technologies that redact or remove personal data.
Governance controls
LinkedIn added an opt-out option for users regarding the use of their data for generative AI training.
The company exempted users in the EU, EEA, Switzerland and the UK from having their data used to train generative AI content models.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.