Faculty and unions are pushing back on how generative AI is governed in education workflows, particularly around evaluation contexts. Reporting emphasizes that teachers can face conflicting or missing state guidance on AI use in portfolios, even as certification outcomes depend on these documents. The St. Paul, Minnesota teachers’ union described contract language aimed at shaping administrators’ AI use in teacher evaluations and potential staff displacement, even as the portfolio debate remains underdefined across jurisdictions. Meanwhile, teacher-preparation leaders say rules may effectively determine employment eligibility because certification tests set the boundaries for permissible work. The policy problem is not only classroom use; it is how institutions operationalize “ethical AI” when certification and employment decisions depend on documentation that administrators may struggle to audit.