Teachers using generative AI to build professional assessment and certification portfolios are running into a patchwork of rules, and labor leaders warn the gaps are producing inconsistent—and potentially unethical—practices. A report highlights how AI can streamline portfolio work that ties evidence to state and district goals, but principals and certification gatekeepers often struggle to identify improper AI use when policies are unclear. The St. Paul, Minnesota teachers’ union president, Leah Van Dassor, said contract language is being used to shape administrator AI use in evaluations while teachers’ own portfolio rules remain unsettled. The article also points to certification tests as a constraint: instructors must coach student-teachers on what is allowed because certification outcomes depend on compliance. Without consistent guidance, the risk shifts from instructional innovation toward academic-integrity disputes within educator credentialing.