San Diego Teacher Integrates AI Into English Classroom

Point Loma High School English teacher Jen Roberts is using classroom-facing AI tools to support reading and writing in her 12th-grade class, according to KPBS. Students read assigned texts for about ten minutes, then interact with Brisk Boost, a tool that shows the source text on one side of the screen and runs a chatbot on the other, prompting comprehension questions tied to Roberts' learning objectives. Scientific American reports Roberts, who has taught for roughly 30 years, also uses MagicSchool, which applies her rubrics and lets students iterate on drafts with immediate feedback; she says the tools return feedback faster than she could alone: "AI doesn't really save me time; it just lets me do more with the time I have."

Reporting describes teachers combining three capabilities: automated rubric-aligned scoring, chatbot-driven formative questioning, and prompt-based idea generation. Scientific American describes a workflow in which a teacher's own grade is compared against an AI "second scorer," with disagreements triggering closer review rather than automatic acceptance. LAist and other coverage flag the standard risk with this approach: generative models can produce fluent but inaccurate output and can reflect biases present in their training data, which complicates their use for grading and assessment.

KPBS reports community pushback alongside the classroom adoption: nearly 1,200 people have signed a petition backing a proposed San Diego Unified School District board resolution that would, among other measures, prohibit student use of generative AI in classrooms. LAist reports California districts are increasingly contracting for AI grading and tutoring tools, while the California Department of Education, per spokesperson Katherine Goyette, does not centrally track which districts are using them. Roberts' classroom sits at the intersection of these two trends: bottom-up teacher adoption for formative feedback, and a state-level oversight gap that leaves data-handling and vendor-vetting decisions to individual districts.

For ed-tech and district technology teams, the operative lesson is that teacher-level AI adoption is outpacing institutional governance: absent centralized state tracking, decisions about vendor contracts, student-data handling, and acceptable-use policy are being made district-by-district and sometimes classroom-by-classroom. That creates inconsistent safeguards even where the underlying pedagogical use, faster and more frequent feedback, is well-supported by teacher testimony.

• •Whether San Diego Unified's school board takes up the proposed resolution restricting generative-AI use in classrooms.
• •Whether the California Department of Education or state legislature moves to establish centralized tracking or guidance for K-12 AI tool adoption.
• •Independent evidence, beyond individual teacher accounts, on whether AI-assisted feedback loops measurably improve student writing or revision outcomes.

Roberts' case is a useful illustration that the practical, defensible uses of classroom AI right now are narrow and teacher-supervised, feedback speed and rubric consistency, not autonomous grading or content generation. The more consequential story is the governance vacuum: when a state education agency cannot say which schools use AI, it also cannot audit data practices or efficacy claims, which is likely to keep fueling parent-led pushback like the San Diego petition regardless of individual teachers' results.