Using Political Power to Protect Students from Using AI in Education

Ever since ChatGPT made its appearance in 2022 and became the most rapidly adopted consumer technology in history, the question of its role in education has been front and center in school circles. From an adult perspective, the impulse to look for ways AI can enhance learning can seem reasonable. I have fruitfully used AI for many purposes myself.

However, I would have no idea if it were being used to positive effect or not if I did not already have a substantial knowledge or experience of the thing that I was using it for. Concerningly, there’s a national movement in education that has great potential to disrupt the learning by which that knowledge and experience gets developed.

Many Americans have never heard of ASCD (Association for Supervision and Curriculum Development) or ISTE (International Society for Technology in Education), but for decades ASCD has been one of the most influential organizations in American public education in setting the agenda for prioritizing major reforms and initiatives in public schools. ISTE has not had quite as broad an influence, but in its area of educational technology it is also the dominant voice that has shaped standards and practice in the use of educational technology across the country.

The two organizations merged in 2023 and have been operating under the somewhat awkward combined brand of ISTE+ASCD. However, in June of 2026, they announced a new organizational name: The International Society for Transforming Education, which sounds ambitious. It’s a somewhat ominous sign for the direction in which they are “transforming education” that one of their first projects after adopting the new name was releasing an expansion of the 2025 ASCD publication: “Profile of an AI-Ready Graduate.”

The vision put forth in the “Profile of an AI-Ready Graduate” affirms that the future makes “uniquely human skills increasingly valuable,” and purports to be about finding balance between “recognizing which skills are uniquely human” and knowing when “it would be problematic” not to use AI. It’s a vision where students “use AI in the service of their goals, while maintaining their own judgment, agency, and creativity.”

There are six broad standards that they propose, which are broken down in more detail in the document:

• LEARNER: Use AI to set learning goals, create plans for learning new skills, identify strategies to get unstuck, and seek targeted feedback to improve performance and understanding.
• PROBLEM-SOLVER: Use AI as a brainstorming partner to clarify challenges, generate new ideas, explore a wide range of possibilities, and evaluate potential solutions.
• RESEARCHER: Use AI to strategically investigate and analyze topics, evaluate claims, and compare sources of information.
• CONNECTOR: Use AI to increase human collaboration, including overcoming language barriers and finding common ground among divergent perspectives.
• SYNTHESIZER: Use AI to synthesize, remix, and refine information into formats and levels of complexity that best meet their unique needs and capabilities.
• STORYTELLER: Use AI to craft compelling narratives and communicate complex ideas through text, image, audio, video, and other media.

There’s a terrific irony in the fact that these areas where ISTE is proposing AI be used are, in fact, “uniquely human skills.” Gaining and developing these capacities is part of how we realize the unique nobility of humanity. And that’s where this set of standards falls on its face. Because it essentially is dictating that students would learn to use AI to support most of their learning functions, completely disregarding the reality that the whole reason students are in school is for them to be taught how to perform those learning functions themselves.

The question that ISTE+ASCD is asking is actually the correct one. The questions of what the role of AI should be in schools, and what we should do about the fact that many students learn to use it badly are vital. But the answers they have come up with are pernicious for learning. I am a high school English teacher in public schools, and as it currently stands, I work hard to prevent my students from using AI in their work. It’s the everyday work where students have to engage deeply with meaningful texts and develop understanding where the actual learning takes place.

However, our technology has now reached a point where AI tools can perform these tasks for the students, potentially "freeing" them from that work and thereby from those opportunities to learn. This set of standards essentially proposes AI as a safety net beneath the students, training wheels to bolt on whenever a certain task feels hard. And it would require that students not just have this available to them, but should be actively taught to turn toward it for all of these purposes. Most of the capabilities that these standards suggest students should learn to use AI to accomplish are functions that we would envision a well-educated person to be proficient in themselves, absent the use of AI. While adopting these standards might produce students who are very capable in their use of AI, it seems to me that it would also produce catastrophic effects on their actual learning of their school material.

As if the potential impact on the efficacy of student learning were not bad enough, the implications for student formation are even worse. In tempting Eve in the garden, the serpent framed the attractiveness of the fruit in terms of how it would make her “like God.” It’s in the difficult wrestling with hard tasks, especially in formative school years, that we learn that we have limits. We build character and discipline by exerting ten more minutes of concentration on a difficult problem when it would be easy to just turn in the work as is and call it “good enough.” It is God who is omnipotent and can create the universe by a mere act of speech; for us mortals, anything worth gaining requires the sweat of effort. However, the ethos ISTE envisions reads to me like a friendly companion proposing, “Why would you work so hard when there’s this nice little AI right here at your fingertips? You’ve made the attempt. Now let the machine finish the job.” AI provides the illusion of omnipotence; it can make you feel like you can do anything, without the effort of the doing, not to mention the learning how to do it. Habituating students to turn to this tool as a first resort strikes me as a disastrous decision in the formation of character and the capacity to be open to the idea that “there is a God, and I am not him.”

And so I was heartened this week when the Florida State Board of Education released a new rule governing the use of AI in K12 schools. The rule is quite long and detailed, but contains such provisions as requiring school districts to adopt an “opt-in” policy for the use of AI in the classroom, where the default is that students do not use it and only parents who specify permitting it for their children have access to it. In the case where access would be permitted, comparable non-AI tools or methods are required to be provided for those who do not opt-in. It requires that AI tools not be used to replace the professional judgment of teachers. It requires that AI tools used in Florida schools be viewpoint neutral and not simulate friendship or an emotional relationship with students.

These are positive points that, in my view, create a healthy friction in the adoption of AI into classrooms, but in my view the most important are that the rules require school districts to “address … permissible classroom usage, verification of student authored work and … demonstration of mastery without the use of AI”, and requiring that any initiative that incorporates AI must be reviewed for effectiveness in “improving student outcomes” i.e. learning.

This is not to say that the new rule is perfect. As may be expected from the highest level of educational leadership in the state, much of its concern is bureaucratic and focused on reporting and privacy compliance. It reflects a tendency to think of the significance of the integrity of student work primarily in terms of assessment, when it's the integrity of the day to day classwork that learning depends on. A provision, for example, forbidding any districts from requiring teachers to assign work where students are expected to use AI would be a substantial improvement. Questions also remain as to how the state would hold districts accountable to comply with this rule.

AI did not become the most rapidly adopted technology in history by being difficult to use. It’s shockingly easy to just log into it and use it for any number of purposes. The idea that schools should orient toward teaching students how to use it for virtually every learning, self-management, and communication function is risible. There is certainly a place for learning to effectively use AI, but the K12 school environment is generally not that place. And AI certainly can be an excellent tool to improve one’s learning, but only after one has gained sufficient knowledge and thinking skill to know what questions to ask and how to evaluate the responses. When schools are fulfilling their function, they are imparting that knowledge and thinking skill to students, and the incorporation of AI on the scale that ISTE+ASCD envisions would be disastrous for their ability to do so. As imperfect as this new rule is, I’m grateful that the Florida Board of Education has seen fit to throw up some roadblocks on the path that so many districts have begun to take. Perhaps in this task they are setting an example for other states to follow.