Show HN: Is this AI training on my prompts? A dated record of the fine print (trainedon.me)

A new compilation of AI tools’ privacy policies reveals that many popular artificial intelligence services are utilizing user prompts for training their models by default, while others require explicit opt-in consent. This comprehensive record documents over a thousand versions of these terms, highlighting significant shifts in how companies handle user data with respect to AI training. Notably, tools like ChatGPT and Claude.ai automatically include user content for training unless users specifically opt out, while others, like Perplexity, strictly prohibit any use of customer content for model training. The findings are crucial as they underscore the ongoing transparency issues within the AI/ML community. Users may inadvertently contribute to the development of these systems without their explicit consent. Changes in policies, such as those announced by OpenAI and GitHub, reflect a growing industry scrutiny over responsible data usage. As AI continues to evolve, understanding these terms is essential not only for compliance and ethical considerations but also for maintaining trust between users and AI providers. This transparency initiative aims to empower users with knowledge about their data rights and prompt greater accountability from AI developers.