HouseCanary's Generative Artificial Intelligence, CanaryAI, is built with a rigorous focus on security and data privacy. Absolutely no data from HouseCanary, including proprietary information, is used to train the large language models (LLMs). Furthermore, if customers use CanaryAI and share any information, all such information is fully protected. It is not used to train the LLM in any way. This ensures that customer data remains entirely confidential and secure within the CanaryAI system.