ChatGPT in the Annotated Bibliography
Northside High School's 11th-grade ELA team teaches a research paper to five sections (140 students). The library media specialist gets three 50-minute sessions with each class. Last year, 22% of annotated bibliographies cited sources that could not be verified, and several students submitted paragraphs that sounded like chatbots. The district has no AI policy. The lead teacher wants to ban phones and AI tools and require handwritten notes. Students have district Chromebooks, and the final paper is 30% of the quarter grade. The librarian has a flexible schedule but is also covering a study hall daily. The principal says any change must be “defensible if a parent complains.” The team meets weekly but the librarian is not on the PLC.
1. How would you redesign the three sessions to make the inquiry process visible and reduce both fabricated sources and undisclosed AI use?
Read the worked answer
I would use the three sessions as checkpoints rather than a tour. Session 1: students frame a narrow question, distinguish source types, and create a shared inquiry log with a required source trail. Session 2: students search the district databases, verify each source's author and date, and complete a short source autopsy in class; I would circulate and conference with each student at least once. Session 3: students synthesize two sources and write a reflection on what changed in their thinking. The teacher would collect the log at each checkpoint, so fabricated citations become visible before the final paper. I would also require students to submit a screenshot or permalink for two sources. This does not eliminate cheating, but it shifts the evidence from the final product to the process. The three sessions would be co-planned with the teacher and tied to the actual paper deadlines.
2. What AI guidance would you recommend in the absence of a district policy, and how would you make it defensible?
Read the worked answer
Without a district policy, I would recommend classroom-level guidance co-created with the ELA team, not a unilateral ban. The guidance would allow AI for brainstorming, grammar feedback, and explaining concepts, require disclosure of any AI use, and prohibit submitting AI-generated text as the student's own. Students would cite AI assistance in a process note and keep drafts. A total ban is hard to enforce, may disadvantage students who follow the rule, and ignores that AI use is already happening. A defensible approach makes expectations public, applies them consistently, and gives students a path to ask questions. I would also document the guidance and send it to parents, so the principal can point to a reasoned process if challenged. If the district later adopts a policy, the classroom guidance can be aligned to it. The goal is academic integrity plus teachable use, not surveillance.
3. How should the team assess process and product so grades reflect learning rather than compliance?
Read the worked answer
The grade should rest on multiple evidence points tied to the standards: question framing, source quality, verification, synthesis, and reflection. I would use a rubric that separates process from product, and I would not give full credit for a polished paper if the inquiry log is missing. Students would complete a short oral defense or conference in which they explain how two sources led to a claim. The annotated bibliography would require permalinks or screenshots, and the reflection would ask students to name a source they rejected and why. The final paper would still carry weight, but process artifacts would be required. This reduces the reward for a single purchased or AI-generated product and gives the teacher evidence to discuss concerns with a student and parent. The team should calibrate the rubric together before grading.