We're excited to share our three SXSW EDU 2027 proposals with you! Community votes help determine which sessions are selected for the conference. Please click the links below to review our proposals and vote for the ones you find most compelling. Voting ends on August 23, 2026, so don't wait! Your support means everything—thank you for helping us bring these important conversations to SXSW EDU!
A National Research Agenda for AI in Education
School systems are making consequential decisions about AI with limited evidence. To address this, researchers are launching a National Research Agenda for AI in Education. This fireside chat introduces the agenda's three high-leverage bets: developing a student's full range of capacities, enabling new educator roles, and building new models of schooling. Speakers will discuss the priority research questions designed to build the capacity to learn from AI adoption while it is underway, showing how districts can help run future studies to test these critical hypotheses.
Session Takeaways
- Takeaway 1: Hear about the three high-leverage bets proposed in the national agenda to guide future research investments and school design choices.
- Takeaway 2: Explore how the agenda plans to study AI's impact on the full range of student capacities, including cognitive and metacognitive skills.
- Takeaway 3: Learn how school districts can participate in the priority research questions to help build a rigorous national evidence base for AI.
Speakers
- Chris Agnew, Managing Director, AI Hub for Education/ Stanford University
- Carly Robinson, Director of Research, SCALE Initiative/Stanford University
Teens, Metacognition, & ChatGPT: Early Lessons from Estonia
When evaluating student-facing AI tools, system leaders must look beyond domestic boundaries to settings where large-scale, empirical data collection is active. This panel shares early findings from an unprecedented nationwide K-12 initiative in Estonia, co-designed by international researchers, technology creators, and state foundations. Rather than focusing on standardized test scores, this research evaluates the metacognitive impacts of ChatGPT use by grade 11 students. Speakers outline the research, teacher perceptions, and early causal findings.
Session Takeaways
- Takeaway 1: Understand how a custom, Socratic ChatGPT tool impacts the everyday learning habits and reasoning of grade 11 high school students.
- Takeaway 2: Evaluate how conversational AI tools shift student metacognitive skills, self-understanding, and critical thinking beyond standard academics.
- Takeaway 3: Discover what nation-level AI research data actually means for a district leader's day-to-day technology strategy and deployment choices.
Speakers
- Chris Agnew, Director, AI for Education Hub, SCALE Initiative at Stanford
- Lily Fesler, Senior Research Associate, SCALE Initiative at Stanford
- Paul Yoo, Senior Research Associate, SCALE Initiative at Stanford
The Evidence Standard: What Actually Works in AI Ed
AI in K-12 is moving faster than the evidence base behind it. This panel cuts through vendor claims to ask: what does rigorous, causal evidence - including randomised controlled trials - actually show about AI's impact on learning? Panelists unpack how to tell whether a tool scaffolds real thinking or shortcuts it, what a diagnostic-plus-human-oversight model can credibly demonstrate, and where the research is still thin. Built for superintendents, teachers, funders, and policymakers who need to separate rigorous evidence from persuasive marketing.
Session Takeaways
- Takeaway 1: How to tell whether an AI tool scaffolds real thinking or trains cognitive shortcuts before signing a contract.
- Takeaway 2: What randomised controlled trials show about diagnostic-plus-human-oversight models in AI learning tools.
- Takeaway 3: How to critically read AI efficacy claims and spot where the evidence is thin or missing entirely.
Speakers
- Chris Agnew, Managing Director, Stanford Accelerator for Learning
- Bibi Groot, Chief Impact Officer, Eedi
- Erin Mote, CEO, InnovateEDU
