Generative AI models like ChatGPT and Bard collect user input, which may include sensitive information. AI services may log IP addresses, browser types, device information, and usage patterns in addition to prompts and chats. Many AI models are developed on massive internet databases that contain personal data collected without consent. This means that an AI's training data may incorporate personal data from internet texts before you interact with it. If you use an AI tool, the provider may store your data on its servers and use it to enhance the model or for other purposes, depending on its privacy policy. Always anticipate that personal data you enter into a public AI service may be collected and avoid revealing confidential information.