The Machine Intelligence Research Institute (MIRI) has submitted a response to the proposed rule from the Bureau of Industry and Security (BIS) for reporting requirements on advanced AI models and computing clusters. We support robust reporting requirements to enhance preparedness for future AI developments and to provide a clear picture of the current state of AI technology and associated risks.We address the quarterly notification schedule, collection and storage of information, and collection thresholds. We suggest implementing ad hoc reporting for significant unanticipated developments and updating the reporting frequency based on the typical length of large AI training runs.Regarding information collection and storage, we emphasize the need for careful handling of sensitive data while ensuring sufficient reporting and oversight. We recommend a tiered system of confidentiality to balance security concerns with the need for information sharing.We highlight that the current thresholds based on computational resources may need adjustment over time due to factors such as algorithmic progress, fine-tuning, and other post-training enhancements. We recommend that BIS regularly update these thresholds based on new information and advancements in the field.Our response provides detailed suggestions for specific information to be requested in the reporting requirements. This includes information about computing clusters (location, usage, security, and emergency response protocols) and AI development practices (capabilities, security measures, and emergency response protocols).