The resulting paper, “ On the Feasibility of Using LLMs to Autonomously Execute Multi-host Network Attacks,” has been cited in multiple industry reports and is already informing safety ...