The Machine Intelligence Research Institute (MIRI) has submitted a response to the RFI for the development of an AI Action Plan, outlining our perspective that AI poses existential risks to humanity. Our response balances innovation with concerns about large-scale risks. We discuss the need for expanded government capacity, maintaining US technological leadership, coordinating with China on AI governance, and restricting dangerous AI proliferation. We emphasize that these recommendations would strengthen US AI governance regardless of one's assessment of extinction risk, providing policymakers with flexible options for addressing coming AI development.