Norwalk schools outline AI framework, classroom pilots and student-facing guidelines

Norwalk — District staff and teachers on Tuesday described classroom pilots of artificial intelligence tools, a draft student-facing AI framework and classroom lessons to teach students how to evaluate AI outputs.

District leaders framed the work as part of the system's Portrait of a Graduate and said the effort aims to help teachers use AI to reduce routine tasks while teaching students how to use the tools responsibly. "We're not telling kids to not use AI," said Martin (presenter). "We're saying to use it responsibly. Talk to your teachers, make sure that you're actually understanding that there's bias in these."

Teacher Caitlin, a classroom teacher at Orkahide, demonstrated a Chrome add-on called Brisk that her high school has piloted. Caitlin showed how the tool can produce a keystroke-by-keystroke timeline for submitted documents, flag large copy-and-paste segments and generate formative quizzes or reading-level–adjusted texts from a transcript. "This tells me that at 08:49 on Monday, March 18, the student had been working on this for 16 minutes," Caitlin said, describing how the tool helps teachers identify when work may not reflect a student's original effort. She also described using the tool to generate Google Forms for quick assessments and to adjust text complexity for diverse learners.

Caitlin and other presenters said the district uses several detectors to check student work for AI generation and runs suspicious items through multiple checkers; the presenters noted that combining tools produces fewer false positives. Caitlin also described classroom approaches to uphold academic integrity, including conversations with students and deliberate test design: faculty sometimes embed unobtrusive prompts in instructions to detect copy-paste from AI, she said.

Mary George, a digital learning coach, demonstrated middle-school uses of generative AI in coding and music, using Code.org's MusicLab Jam Session to show how students explore pattern generation and the "temperature" parameter that makes AI output more or less adventurous. "They're learning these concepts of AI," Mary George said, describing how block-based coding now integrates AI features that students can experiment with.

Presenters said the district has an AI framework developed with an advisory committee and is drafting a student-facing version and a version for younger students; staff expect to circulate those materials to schools in coming weeks. Martin said the district is preparing infographics and posters in student language on topics such as responsible use, safety and ethical considerations.

Presenters noted cost implications: the classroom demonstrations relied in part on paid tiers of some tools, and staff said the district may need to consider licensing for schoolwide deployment. "For a lot of the tools that we would need, we really need the paid version," Caitlin said.

District officials also discussed classroom lessons on AI bias, showing an exercise adapted from Common Sense Media that had students compare outputs and identify training-data limitations. The lesson asked students to evaluate when an AI-generated image or text might reflect underrepresentation or stereotypes and to suggest ways to reduce bias in training data.

The workshop ended with staff encouraging continued work on student-facing guidelines, teacher training and selection of platforms so schools can share experience and scale promising practices.