In this edition
College professors are reviving the oral exam
NYC released its first AI guidance for schools
Boston is bringing AI literacy to every public high schooler
A Wichita teacher is using AI to make reading engaging
College professors are reviving the oral exam
Professors at Cornell, Penn, and NYU are reaching for one of the oldest assessment tools there is: sit down and explain it to me. Cornell students sign up for 20-minute Socratic defenses after submitting problem sets. Their written work is no longer graded at all, just the defense. At Penn, a professor now pairs oral exams with written papers, not to catch cheating, but because she wants students to actually build their own critical thinking capacity. NYU's Panos Ipeirotis went a step further and built an AI-powered oral exam using a cloned professor's voice (via the AI tool ElevenLabs). It adapts its questions in real time based on each student's answers, drilling deeper where answers are thin and offering clues when they stumble. It's designed to verify that students actually understood and engaged with what they submitted. AI didn't create the problem of students not owning their thinking. It just made the problem impossible to ignore.
Sponsored by Legend.org -- The AI assistant built for the world’s best teachers. Personalized insights, longitudinal student analytics, and visibility into what to focus on next.
NYC released its first AI guidance for schools
New York City released its preliminary AI guidance this week, offering a traffic light framework for 78,000 teachers: green for lesson planning and brainstorming, yellow for student data and translations with human review, and red for a defined set of cases the city considers off limits, including IEPs, grading, and counseling. The framework is intentionally high level, some of the most contested questions: like how students can use AI for homework and how teachers can use it to support feedback and IEP differentiation are still unresolved. A public comment period runs through May 8, with a fuller playbook expected in June.
Boston is bringing AI literacy to every public high schooler
The same week NYC released its framework, Boston moved in a different direction. Mayor Michelle Wu announced a partnership with UMass Boston, backed by a $1 million contribution from tech entrepreneur Paul English, to bring AI literacy to every public high school student starting next school year. The curriculum is being built this summer, focused on ethics, creative thinking, and AI literacy rather than just technical skills. Students will also have access to credit-bearing AI courses at UMass Boston. Boston is asking what students need to learn. NYC is asking what the guardrails should be. Both questions need answering.
A teacher is using AI to make reading engaging
A special ed teacher in Wichita described a colleague who started using ChatGPT to generate reading passages on topics his students actually care about, from dinosaurs to Power Rangers to Lionel Messi. The result: students requesting more reading once they're done with their work. It's a small example but it points at something important. Engagement might be the most underrated variable in learning. When a student genuinely wants to learn something, everything changes, and we can all relate. AI can help on two fronts here: making material uniquely interesting for each student, and freeing teachers from time-consuming work so they can focus on what actually moves the needle getting students curious, excited, and genuinely engaged with the work.
Lots moving in AI education this week: policy, frameworks, mandates, the White House’s AI Summit... But the story I keep coming back to is the one about kids asking for more reading. If you come across a story worth sharing, reply and let me know.
- Matt
Prompt & Response: AI and education news worth thinking about. Join 2,000+ readers by subscribing below: