This report discusses strengths and weaknesses of using current large language model (LLM) tools for public comment processing. It draws on lessons learned from applying an LLM tool to support the analysis of responses to the public Request for Information