Episode 87

What EdTech Gets Wrong About Implementing AI

Guest name: Stephen Jull
Guest title: Global Head of AI & EdTech
Guest company: Teach For All
Host: Sean Strathy

Brief description of the episode

Before your next AI strategy conversation, ask yourself this: are you starting with implementation or readiness? Stephen Jull, Global Head of AI and EdTech at Teach For All, argues that schools may be starting in the wrong place. The real question is not just how to roll out AI, but whether the institution is ready for it.

In this Tech in EdTech episode, Stephen joins Sean Strathy to discuss why AI readiness must come before rollout, and what schools need to get right before making their next major AI decision.

Key Takeaways:

- Assess educator readiness before rushing into tool selection or implementation.
- Help teachers clarify their pedagogical goals, content knowledge, and motivations before asking them to use AI meaningfully.
- Move beyond basic tool training and support educators in understanding how AI fits into who they are as teachers and what they are trying to achieve.
- Treat readiness as the starting point for AI strategy, not something added after procurement.
- Recognize that AI takes the shape of the person using it, which makes educator judgment, clarity, and context essential.
- Re-examine whether the product solves a real learning problem or simply adds AI to an old workflow.
- Recognize that foundation models can already help teachers differentiate content, adjust reading levels, and create learning supports that once required third-party tools.
- Shift product value from making educators dependent to making educators more capable.
- Build AI features that strengthen teacher oversight instead of isolating students inside standalone AI interactions.
- Focus on problems that become more urgent in an AI-enabled classroom, including assessment, responsible student use, equity of access, and evidence of learning.
- Stop applying the old edtech rollout model to AI. The usual procurement, deployment, and professional development cycle was not designed for a general-purpose technology.
- Avoid assuming there is one correct model, platform, or implementation path for every classroom or system.
- Give educators the freedom to experiment responsibly with foundation models before locking into rigid top-down decisions.
- Build an implementation strategy from what teachers are actually discovering in real classroom contexts.
- Share what is working across schools, networks, and systems so institutions can learn from each other as the technology evolves.
- Recognize that schools can no longer rely on being the main gatekeepers of knowledge.
- Shift the institution’s value from knowledge transfer alone to helping learners understand, question, apply, and connect knowledge.
- Invest in pedagogy over content delivery, because the educator’s role becomes more about guiding each learner’s journey.
- Help students build the judgment to ask better questions and make sense of information, not just retrieve it.
- Redesign learning experiences so students analyze, synthesize, apply, and create rather than only recall.
- Teach For All: https://teachforall.org/
- Stephen Jull’s session with Reed Hastings, on “Teaching in the Age of AI”:https://youtu.be/aW6xoMaaQFs?si=bHvABHBxTSQb-cKy
- Further reading: Before You Implement AI in Education, Build the Guardrails

What EdTech Gets Wrong About Implementing AI

Brief description of the episode

Key Takeaways:

You may also like this

Get the Citation‑First AI Checklist