From Frameworks to Classrooms - Closing the AI Literacy Implementation Gap
AI literacy frameworks already exist. Who turns them into lessons, what support do teachers need, and how can schools assess what students learn? An online panel brought education, policy and practical AI literacy work together as part of Catalyst Now Week 2026.

The question
An AI literacy framework can set out what students should learn. Turning it into something a teacher can use takes more work.
On 16 October 2026, Young AI Leaders Linz organized an online panel during Catalyst Now Week to discuss how frameworks such as DigComp 3.0 and the EU–OECD AI Literacy Framework could become part of everyday teaching.
A recurring concern was the difference in pace. Students and families are already making decisions about AI, while education systems are still working out what to teach, who should teach it and how to make it accessible.
The panel
The panel brought together Laura Knight, founder of Sapio, author and AI in education advisor, and Jan-Willem van Putten, Co-Founder and Senior Program Manager for AI in Europe at The School for Moral Ambition.
They were joined by Dino Osmanagić, Hub Leader of Young AI Leaders Linz and Head of Innovation & Product at incert, and John Ofori, Founder and CEO of Rethink Africa Intelligence and a member of Ghana's National AI Strategy Working Group.
Andreas Patuzzi hosted and moderated the session, with Charles Walcher as co-host.
Policy and responsibility
The panel argued that national education policy needs a closer connection to what happens in schools. Including AI literacy in a curriculum still leaves teachers with questions about how to teach it, which materials to use and where to find support. Teachers should be able to help shape policy and report which approaches work in their classrooms.
Schools already trying new approaches can offer examples for others to test and adapt. The discussion made a case for combining national guidance with this local experience, while checking the quality of what is taught and sharing useful practice.
Teacher training needs support of its own: qualified trainers, clear accreditation and someone responsible for keeping standards up to date. Other responsibilities were less settled. Who should monitor generative AI use, and who should provide the safeguards? The discussion raised these questions without resolving them.
Teaching AI literacy
One practical proposal was to teach through realistic scenarios. Students could examine an AI-generated answer, check its claims, consider its limitations and explain whether they would rely on it. The exercise gives them time to reflect and practice judgment, rather than stopping at a quick reaction to whether AI is good or bad.
Teachers need experience of this themselves. The panel called for AI literacy to be included in teacher education and discussed routine tasks as a possible starting point. Trying AI on work they find burdensome could help teachers understand its uses and limitations without treating it as another subject to fit into an already full timetable.
The discussion also considered what AI means for homework, exams and established ways of teaching. Curious students have more ways to explore a subject, but schools still need to see what they understand. The panel saw room to keep teaching methods that work and reconsider those that no longer serve learning.
Unequal access
Students also arrive with very different support. Some have schools or tech-savvy parents helping them use AI revision tools. Others are told not to use AI at home, cannot afford a subscription or lack reliable internet. A lesson that assumes everyone can use the same tool may widen an existing learning gap.
Participants also questioned subscription prices that take little account of differences in purchasing power across countries. Examples from Africa drew attention to infrastructure barriers and difficulties aligning institutions and schools on what to teach, particularly in rural areas. These concerns affect which teaching approaches a school can realistically offer and which students can benefit from them.
Assessing learning
The panel cautioned against trying tools simply because they are new. Schools need to decide what they want to improve before choosing a tool or judging its results. Even a claim that AI saves ten hours says little about educational value if that time is immediately filled with other work.
The same question applies to students' work. An AI-assisted homework submission may tell a teacher little about the student's own understanding. Live discussion, practical tasks and some work with pen and paper were suggested as ways to see how students reason and reach decisions.
Participants proposed surveys after training and performance-based tests to assess whether AI literacy programs improve skills. Some gains will be harder to count than time saved, but they should be visible in how students question an answer, explain their thinking and decide what to trust.
On stage & involved
- PanelFounder · Sapio
- PanelCo-Founder and Senior Program Manager for AI in Europe · The School for Moral Ambition
- PanelCEO · Rethink Africa Intelligence
- PanelHub Leader · Young AI Leaders Linz
- ModerationRegional Lead · Young AI Leaders Linz
- ModerationDeputy Team Lead · Young AI Leaders Linz
Part of Catalyst Now Week 2026
- Catalyst Now Week 2026 — Catalyst Now
No press release for this one yet. Writing about it? Reach our press desk via the contact page or grab logos and boilerplates in the media kit.
Share that you're thereMake your own social card with your photo and name — it stays on your device.Open card makerClose
Pick your role first — speakers, volunteers, attendees and partners each get their own design and post text.
No photos published for this event yet.
Want to build the next one? Join the hub