Another major tech company is using user data to train AI models by default, without prior user notification. Following Meta and X’s Grok, LinkedIn is automatically enrolling users into its AI training, including models from unnamed affiliates.
Owned by Microsoft, LinkedIn clarified that while user data won’t train base OpenAI models, it will be shared with Microsoft for its own AI software. LinkedIn stated that its generative AI features might be trained by LinkedIn or other providers, such as Microsoft’s Azure OpenAI service.
Greg Snapper, a spokesperson, mentioned that LinkedIn uses OpenAI models via Microsoft’s Azure AI Service, without sending data back to OpenAI for training.
Mariano delli Santi from the Open Rights Group criticized the opt-out model, emphasizing the need for opt-in consent to protect user rights. LinkedIn assured that it minimizes personal data in training datasets and does not train content-generating AI models on data from the EU, EEA, or Switzerland.
Users can opt out by adjusting their data privacy settings. However, privacy advocates remain concerned about LinkedIn’s decision to opt users into multiple AI training models, urging regulatory action against such practices.
