The use of generative artificial intelligence (GAI) large language models has increased in both professional and classroom technical writing settings. One common response to student use of GAI is to increase surveillance, incorporating plagiarism detection services or banning certain composing activities from the classroom. This paper argues such measures are harmful and instead proposes a "CARE" framework: critical, authorial, rhetorical, and educational--a nuanced approach emphasizing ethical and contextual AI use in technical writing classrooms. This framework aligns with plagiarism best practices, initially devised from when rhetoric and composition scholars considered the pedagogical implications of the Internet.