The AI Observatory now aggregates consented conversations from seven datasets to analyze how people actually use generative AI. This public platform bypasses curated reports from OpenAI and Anthropic. It provides the first independent corroboration of user behavior. Researchers can now verify if model capabilities match real-world application patterns.