Consent in Training AI

LinkedIn and Udemy have recently been using user personal data for training generative AI models without properly notifying users or updating their privacy policies. LinkedIn has an opt-out toggle for users, but Udemy's opt-out window has already closed, leaving users with no choice. This raises ethical concerns about data privacy, risk, and informed consent. Personal data creates risk when used for training AI, as it increases the chance of data breaches and unintended exposure. The principle of informed consent should be taken seriously, and dark patterns in user experience design can be coercive. This issue also applies to images and multimedia, not just text. Affected users have limited options for reacting to these practices, but can opt out if allowed or seek regulatory action under data privacy laws like CCPA.