By Chuck Gallagher | Business Ethics Keynote Speaker | AI Speaker and Author
TL;DR: Chuck Gallagher, business ethics keynote speaker, argues that the 2026 fight over AI in classrooms will not be won by the slickest app, but by the schools that treat AI adoption as an ethics decision before a technology decision. Drawing on Emily Freitag’s recent Fordham Institute commentary, the real question for superintendents and principals is not which tool to buy, but who is accountable for what those tools do to students, teachers, and learning itself.
A teacher in a mid-sized district recently told me she has eleven different AI tools open on her browser before her first bell rings. Eleven. None of them were chosen by her principal, none were vetted by her district, and none came with a single sentence about what data they collect on her students. She is not unusual. She is the norm. And that norm is the quiet ethics crisis that nobody in the edtech conversation wants to name.
Emily Freitag, the cofounder and CEO of Instruction Partners, recently published a set of predictions for AI in education in 2026 through the Thomas B. Fordham Institute. Her central claim is that student learning gains in the coming year will correlate more with the quality and specificity of instructional leadership and teacher support than with which product a school happens to buy. I think she is right. I also think she is identifying an ethics problem dressed up in product-management language.
Who Is Actually Accountable When the Algorithm Teaches?
Freitag predicts that the early surge in teacher-autonomy AI tools, the kind individual teachers download with freemium accounts, will start losing market share in 2026 to products designed for district-wide use. The reason she gives is practical: it is harder to prove impact when every teacher uses a different tool in a different way. As a business ethics keynote speaker, I would add a harder reason. When eleven different tools are running in eleven different classrooms, no one is accountable for the cumulative effect on a child. Not the teacher who downloaded them in good faith. Not the principal who never saw them. Not the vendor whose terms of service nobody read. Diffused responsibility is how almost every organizational ethics failure I have studied actually happens.
This pattern is not new. The OECD Digital Education Outlook 2026 warns that overreliance on AI for tasks like marking, feedback, and lesson planning can erode teachers’ professional judgment when systems operate opaquely. A 2026 conceptual study published in Education Sciences put it more bluntly, describing the risk of “pedagogical deskilling” when generative AI displaces the diagnostic and reflective work through which teachers build expertise. Translation: when a tool quietly takes over the thinking, the human eventually forgets how to do it. And once the humans forget, the institution has nothing left to fall back on when the tool fails.
Choosing a Tool Is Choosing a Set of Values
Freitag draws another important line in her piece. Some AI products are built to “meet students where they are,” offering features like text releveling that drop content down to a child’s current reading level. Other products anchor to grade-level instruction and assume every student can be supported up to that bar. These look like product-design choices. They are actually moral choices about what we believe children are capable of, and what we owe them. A district that adopts the first kind of tool at scale is making a quiet bet that some kids will be permanently met where they are. That is not a software preference. That is a values statement, whether the school board ever votes on it or not.
Throughout my career talking with leaders about choices and consequences, I have watched the same pattern play out in finance, healthcare, and now education. People mistake a procurement decision for a neutral act. It is never neutral. Every tool encodes assumptions about who matters, who decides, and what counts as success. The HolonIQ market projection puts AI in education at $12.3 billion globally by 2026. That is a lot of value flowing into products whose ethical assumptions almost no school board has formally examined.
What Leaders Should Actually Do in 2026
Freitag closes her predictions by noting that her track record on forecasts is wrong about as often as it is right. I appreciate that honesty, because the temptation in this space is to sound certain. As an AI ethics speaker and author, the recommendation I would put in front of every superintendent and principal in 2026 is unglamorous: before you buy or expand a single AI product next year, write down on one page who is accountable for which outcomes, what data leaves your building, and what you will do when the tool gets something materially wrong. If you cannot answer those three questions in plain English, you are not ready to scale the tool. You are ready to scale the risk.
I have argued at ChuckGallagher.com for years that the cybersecurity failures, the leadership scandals, and the institutional collapses that dominate the headlines are almost never technology failures at the root. They are ethics failures wearing a technology costume. AI in classrooms will follow the same script unless leaders refuse to outsource the moral weight of the decision to the vendor. The schools that thrive in 2026 will be the ones whose leaders treat AI adoption as a question of accountability first, capability second.
Frequently Asked Questions
What did Emily Freitag predict about AI in education for 2026?
Emily Freitag, cofounder and CEO of Instruction Partners, predicted in a Fordham Institute commentary that 2026 student learning gains will depend more on the quality of instructional leadership and teacher support than on which AI product a school selects. She also forecast that teacher-autonomy AI tools will lose ground to products built for district-wide implementation, because impact is easier to prove when use is consistent.
Why is AI adoption in schools an ethics issue and not just a technology issue?
AI adoption in schools is an ethics issue because every tool encodes assumptions about who is accountable for student outcomes, what student data is shared, and what beliefs the system holds about what children can achieve. Chuck Gallagher, business ethics keynote speaker, argues that diffused accountability across many uncoordinated AI tools is the structural condition under which most organizational ethics failures occur, in classrooms or anywhere else.
What is pedagogical deskilling and why does it matter for AI in classrooms?
Pedagogical deskilling is the gradual loss of teacher expertise when AI systems take over diagnostic, reflective, and decision-making tasks that teachers previously performed themselves. A 2026 conceptual paper in Education Sciences warned that opaque or autonomous AI tools can displace the cognitive work through which teaching expertise is built and sustained. The OECD Digital Education Outlook 2026 raised the same concern about marking, feedback, and lesson planning being heavily automated.
How big is the AI in education market expected to be in 2026?
HolonIQ projects the global AI in education market at approximately $12.3 billion in 2026, with a roughly 36 percent compound annual growth rate since 2022. That growth means the volume of AI-driven decisions affecting students will increase sharply in the coming year, even before most districts have formal accountability frameworks for those decisions in place.
What should school leaders do before adopting an AI tool district-wide?
Before adopting an AI tool district-wide, school leaders should be able to answer three questions in plain language: who is accountable for which student outcomes, what data leaves the building when the tool runs, and what the response plan is when the tool gets something materially wrong. If a district cannot answer those three questions clearly, it is not yet ready to scale the tool. It is only ready to scale the risk.
Join the Conversation
I would genuinely like to hear from teachers, principals, superintendents, and parents on this one. If your district has adopted an AI tool in the last twelve months, what is the one accountability question you wish someone had asked before the contract was signed? Drop your answer in the comments below. The questions that follow are designed to give you and your team a starting point for a more honest conversation about what you are actually choosing when you choose an AI product for children.
Five Questions for Further Thought and Consideration
- If a student is harmed by a recommendation an AI tool made in your school next year, who in your organization will answer for it, and have they been told yet?
- Which AI tools are currently in use in your classrooms that were never formally approved by district leadership, and what does that gap say about how decisions actually get made?
- When your AI vendor changes its terms of service, its data handling, or its underlying model, what is your process for re-evaluating whether the tool still meets your ethical standards?
- If a parent asked you tomorrow exactly what AI is doing with their child’s work, their performance data, and their behavioral patterns, could you answer in language they would understand?
- What are you willing to give up, in cost savings or convenience, to keep human judgment at the center of decisions that meaningfully affect a child’s academic future?
Related Articles:
Cyber Security Report 2026: AI Is Now an Ethics Problem
AI Agents Are Making Decisions for Your Business. Who’s Accountable?
