This is a Khan Academy product/demo webinar about new interim assessments, focused on moving beyond multiple-choice toward more authentic, AI-assisted, and actionable assessment experiences. Lauren Deeders explains the product rationale and theory of action; Peter Jacobson demos the student experience, explain-your-thinking conversations, accessibility features, scoring architecture, and future personalized recommendations.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This transcript is a product launch / walkthrough for Khan Academy’s new interim assessments, framed as part of a “re-imagined” Khan Academy. The core thesis is that traditional assessments are too multiple-choice heavy, too slow to produce useful action, too opaque to build trust, and too weak at capturing real student thinking. Lauren Deeders argues that Khan Academy is using AI plus psychometrics to build a more authentic, efficient, and open assessment system that can better measure what students know and can do. Lauren’s setup is that Khan Academy already has formative tools like unit tests, quizzes, and course challenges, but interim assessments fill a different need: a more comprehensive assessment and learning system. …
Immediate setup is product validation: watch whether the explain-your-thinking flow feels natural, fast, and non-punitive in early district use. The main near-term risk is friction from mandatory responses and any sign the AI prompts students too much or too little.
Over the next few months, the base case is a phased rollout in math first, with ELA still being tuned and district alignment doing most of the work. The key confirmation signal will be whether teacher-facing summaries translate into actionable instruction and repeat usage.
The long-run thesis is a shift toward assessment as an always-on learning system rather than a closed testing event. If the model works, it weakens the old tradeoff between rigor, transparency, and actionability by using AI to surface reasoning at scale.
Khan Academy is building brand new interim assessments as part of Re-Imagined Khan Academy.
The presenters repeatedly frame the session as a preview of new interim assessments in the reimagined product.
The company believes multiple-choice-only assessment is insufficient because it does not capture real-world application or nuanced understanding.
Lauren says multiple choice is efficient but too binary and not representative of real use.
Conversational assessment can mimic the kinds of teacher-student probing used to understand student thinking.
Lauren explicitly says AI can prompt and probe like a teacher asking follow-up questions.
Why are you developing interim assessments, and why now?
Lauren explains that Khan Academy already has formative assessments like unit tests and quizzes, but they saw an opportunity with AI to create a comprehensive assessment and learning system for interim assessments. They identified pain points including: too much multiple-choice, lack of actionability, lack of trust (guarded item banks), and assessments that don't fit in a class period. They aim to address these with authentic assessments (interactive items, short response, conversational assessments), actionable score reports, open item banks, and efficiency.
How does the explain your thinking feature work in Khan Academy's assessments?
Peter walks through a demo of the explain your thinking feature. It starts with a two-part question where the student first answers a math or ELA item using standard input widgets, then enters a conversation with an AI persona called Conductor. The AI probes the student's conceptual understanding (e.g., asking why the tangent of 30.5° is the same in all right triangles with that angle). If the student demonstrates sufficient understanding, Conductor ends the conversation. If not, it continues probing. The system uses three AI agents: one to score the response, one to generate the conversation, and a self-critique agent that checks the other AI's work to ensure it doesn't give away the answer.
What kinds of item types are available in Khan Academy's new assessment experience?
Peter shows several item types including numerical input, geometry proofs using a drop-down format, expression widgets with a number pad for complex equations, multi-check or single-select multiple choice, the explain your thinking conversational AI item, short response items, and interactive graph items that are accessible via keypad for screen readers.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
