For the past few years, the dominant conversation about AI and education has revolved around a single anxious question: what happens when students use AI to cheat? It was a reasonable concern and it still is. But in June 2026, that conversation is finally being joined by a more constructive one. How do we actually use AI to help students learn better, and what does that look like in practice at scale?
Two pieces of research and one major product announcement this week are pushing that second conversation forward in ways that are hard to ignore.
Microsoft Releases Its Third Annual AI in Education Report
On June 24, 2026, Microsoft published the third edition of its annual AI in Education Report, and the headline is that adoption has moved well past the experimentation phase. Alongside the report, the company announced a new set of AI-powered teaching and learning tools that will be available at no additional cost to Microsoft 365 Copilot users.
The most significant of these is Copilot Notebooks, a workspace that allows students to focus their learning around their own course materials. Rather than leaving students to navigate AI in an open-ended way, the feature is designed to keep AI assistance tethered to what students are actually studying. The idea is to reduce the risk of AI producing generic or irrelevant output and increase the likelihood that the time students spend with it is genuinely productive.
Alongside Copilot Notebooks, Microsoft announced the Study and Learn Agent, which brings research-based learning into Copilot Chat. The agent guides students through concepts using interactive practice and real-time feedback, drawing on principles from educational psychology about how people retain information most effectively. It is a more structured approach than simply asking a chatbot for help, and that structure is deliberate.
What the OECD Found About GenAI and Teaching
The OECD published its Digital Education Outlook 2026 this month, and the central finding is nuanced in a way that both optimists and skeptics about AI in education should take seriously. The research suggests that generative AI can genuinely support learning, but only when it is guided by clear teaching principles and when teachers are actively involved in shaping how it is used.
The OECD’s conclusion is that the most effective outcomes occur when teacher expertise is integrated into the design of AI tools from the start, rather than added as an afterthought. When that happens, generative AI can amplify what teachers are able to do, creating benefits that neither teachers working alone nor AI working alone could achieve. That finding has significant implications for how schools and districts should be evaluating and adopting AI tools. The question is not simply whether a tool is technically capable, but whether teachers have genuine input into how it works.
Lawmakers Are Paying Attention
The pace of adoption is drawing legislative scrutiny. FutureEd’s 2026 State AI in Education Legislative Tracker is currently monitoring 71 bills across 27 states, all of which address some aspect of artificial intelligence in classroom instruction. The breadth of proposed legislation reflects genuine uncertainty about how AI should be governed in schools.
Some states are moving toward relatively permissive frameworks that prioritize experimentation. Others are taking a harder line. South Carolina’s H.B. 5253, for example, would require written parental opt-in consent before AI tools can be used with students, prohibit AI from replacing licensed teachers in core instruction, and ban AI-driven high-stakes decisions about students without human oversight. Whether or not that particular bill passes, it signals the direction in which parent and community concern is pointing.
From Experiment to Governance
The clearest theme running through all of this week’s education AI news is a shift in the fundamental challenge facing schools and policymakers. The question is no longer whether to use AI. That decision has largely been made, in practice if not always in policy. The question now is how to govern it, which tools deserve trust, what guardrails protect students, and how to ensure that the humans in the process remain genuinely in control of the most important decisions.
That is a harder and more interesting question than the cheating debate that consumed so much energy in 2023 and 2024. And the fact that governments, international organizations, and major technology companies are all wrestling with it seriously at the same time suggests that education is entering a more mature phase of its relationship with artificial intelligence.
The tools are arriving faster than the frameworks to govern them. The work of 2026, for educators and legislators alike, is to close that gap before it widens further.
For the latest on AI across technology, health, education, and more, explore Mylistingo.
