From AI Skeptic to AI Champion, Part 2

This article is part of the series “From AI Skeptic to AI Champion,” which explores how educators can move from fear and uncertainty toward wise, ethical, and human-centered AI adoption.

Since ChatGPT reached 100 million users within two months of launch, educators have been scrambling to keep up with students who adopted the technology almost immediately (Hu, 2023). For many teachers, the first response was not excitement. It was fear.

This fear was not irrational. Generative AI raised real concerns about cheating, safety, privacy, student dependency, misinformation, bias, assessment, and even the future of the teaching profession. Teachers and school leaders were right to pause and ask hard questions. But fear, while understandable, cannot be our final answer.

Fear can alert us to danger. It can slow us down when we are moving too quickly. It can keep us from being naive. But fear is not enough to build wise policy, healthy school culture, or effective teaching practice. If fear becomes the dominant posture, we will either ban what we do not understand, ignore what we cannot control, or police students in ways that damage trust. In my work with educators and schools, I have seen three great GenAI fears emerge again and again.

Fear #1: AI Means Cheating

Many educators have equated AI with cheating. This is the fear that arrived first and loudest. Teachers quickly realized that students could use ChatGPT to write essays, answer discussion questions, summarize readings, complete assignments, and produce work that appeared polished and original. The concern was obvious. If AI can do the work for students, how will we know what students actually know? And are students actually learning?

This is a serious question. We should not trivialize it. Academic integrity matters. Student thinking matters. Writing matters. Practice matters. Struggle matters. Learning often requires effort before it produces fluency. At the same time, our first responses to this fear were not always helpful.

Some educators turned immediately to AI-detection software, hoping for a technological solution to a technological problem. But detection tools have proven unreliable. They can produce false positives, wrongly accusing students of misconduct (Liang et al., 2023; Perkins et al., 2024). They can also be bypassed, revised around, or manipulated through simple prompting and editing strategies.

When teachers rely too heavily on detection, the teacher’s role begins to shift. The teacher becomes police officer, prosecutor, judge, and jailor. Students become suspects. The classroom becomes a place of surveillance rather than formation. We need a better approach.

A peer-reviewed study by Stanford education scholars Victor R. Lee and Denise Pope, along with Sarah Miles and Rosalia C. Zarate, complicates the assumption that ChatGPT suddenly caused a cheating crisis. In their study of anonymous survey data from three high schools before and after the public release of ChatGPT, the researchers found that self-reported cheating behaviors remained relatively stable after generative AI tools became widely available (Lee et al., 2024). The study does not suggest that AI misuse is harmless or imaginary. Rather, it reminds us that academic dishonesty is not new, and that students’ reasons for cheating often reach deeper than access to a particular technology. In other words, AI did not create the temptation to avoid learning; it gave students a new tool for an old temptation.

The answer cannot be only catch and punish. The better answer is to teach, to model, to form, to cultivate, to nurture. We need to teach students what appropriate AI use looks like. We need to redesign some assessments. We need to evaluate process, not only product. We need to use in-class writing, oral defense, reflection, revision histories, student conferences, and assignments that require personal application, local context, and authentic thinking. Most importantly, we need to help students understand why learning still matters when answers are easy to generate.

The question is not, “Did you use AI?” The better question is “How?”

Fear #2: AI Is Dangerous

The fear that AI is unsafe is a legitimate one. AI tools can produce misinformation. They can hallucinate. They can reinforce bias. They can mishandle sensitive information. They can expose students to inappropriate content if tools are not designed or supervised well. They can give confident answers that are simply wrong. And AI can be used by humans to do harm. For younger students especially, safety must be taken seriously. But here again, avoidance is not the same as wisdom.

Some schools initially responded by banning AI. New York City’s Department of Education and Los Angeles Unified School District were among the large districts that initially blocked ChatGPT. Eventually, many districts reconsidered. By early 2024, Education Week reported that 73% of educators surveyed said their districts did not prohibit generative AI tools, 20% reported that teachers were allowed to use the technology but students were barred from using it, and 7% reported that both teachers and students were banned from using generative AI in school (Klein, 2024).

This movement from prohibition toward guided use reflects a growing recognition: the most dangerous approach to AI may be having no approach at all. If schools ban AI without teaching students about it, students will still encounter it elsewhere. They will use it at home, on personal devices, through search engines, in social media, inside productivity tools, and eventually in the workplace. A school that refuses to discuss AI does not protect students from AI. It simply removes educators from the conversation. The better path is guided use.

Schools need age-appropriate tools, clear policies, teacher training, privacy protections, parent communication, and wise supervision. Tools such as Khanmigo have attempted to address some of these concerns by building in student safety guardrails, privacy protections, teacher transparency, and parent visibility (Khan, 2024). These features do not eliminate all risk, but they demonstrate an important principle: AI in education must be designed and governed differently than AI for casual adult use. 

Students should not be left alone to figure this out. They need adults who can help them ask: Is this true? Is this biased? Is this ethical? Is this helping me learn, or is it helping me avoid learning? What information should I never share with AI? When should I verify? When should I cite? When should I turn the tool off? 

Safety is not achieved by silence. Safety is achieved through wisdom, understanding, structure, practice, engagement, conversation, struggle, and honesty.

Fear #3: AI Will Replace Teachers

The third fear is the most personal. Teachers worry that AI will replace them. As GenAI capabilities accelerate, this fear has become more common. AI can already generate lesson plans, produce quizzes, explain concepts, write feedback, summarize texts, translate languages, and tutor students conversationally. It is not difficult to imagine administrators, policymakers, or technology companies becoming overly enamored with efficiency and underestimating the human complexity of teaching.

AI will change teaching. But AI cannot replace the fullness of a teacher’s vocation. Teachers are not merely content deliverers. Teachers form students. Teachers build trust. Teachers notice tears, boredom, confusion, anxiety, pride, avoidance, and joy. Teachers know when to push and when to pause. Teachers understand the difference between a wrong answer and a wounded student. Teachers create communities of learning where students are seen, known, challenged, and loved.

At the same time, teachers who adopt and integrate AI faithfully and effectively may be better positioned to serve students than those who refuse to engage it at all. That is not because AI is better than teachers. It is because wise teachers can use AI to reduce repetitive burdens and return more fully to the work only teachers can do.

This is the paradox. AI may automate some tasks, but it can also restore the teacher’s attention to the more human work of teaching. If AI can help draft a rubric, the teacher can spend more time conferencing with students. If AI can help create differentiated materials, the teacher can spend more time guiding small groups. If AI can help generate practice questions, the teacher can spend more time listening to how students think. If AI can help with administrative communication, the teacher can spend more time building relationships.

The goal is not to make teachers less necessary. The goal is to free teachers to be more fully present.

From Fear to Faithful Practice

These three fears are real: cheating, danger, and replacement. But fear must be transformed into wisdom. With cheating, we move from detection to formation, assessment redesign, and academic integrity. With safety, we move from avoidance to supervised, age-appropriate, transparent use. With replacement, we move from professional anxiety to vocational clarity. The teacher’s role is not disappearing. But it is changing. And if teachers are going to lead students well in an AI-shaped world, we must become learners again. We must read, watch, listen, experiment, question, verify, and reflect. We must be cautious, but not closed. We must be curious, but not careless. We must help students understand not only how to use AI, but when to use it, why to use it, and when not to use it at all.

References

Hu, K. (2023, February 2). ChatGPT sets record for fastest-growing user base – analyst note. Reuters. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/

Khan, S. (2024). Brave new words: How AI will revolutionize education and why that’s a good thing. Viking.

Klein, A. (2024, February 28). Does your district ban ChatGPT? Here’s what educators told us. Education Week. https://www.edweek.org/technology/does-your-district-ban-chatgpt-heres-what-educators-told-us/2024/02

Lee, V. R., Pope, D., Miles, S., & Zarate, R. C. (2024). Cheating in the age of generative AI: A high school survey study of cheating behaviors before and after the release of ChatGPT. Computers and Education: Artificial Intelligence, 7, 100253. https://doi.org/10.1016/j.caeai.2024.100253

Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), 100779. https://doi.org/10.1016/j.patter.2023.100779

Perkins, M., Roe, J., Vu, B. H., Postma, D., Hickerson, D., McGaughran, J., & Khuat, H. Q. (2024). GenAI detection tools, adversarial techniques and implications for inclusivity in higher education. International Journal for Educational Integrity, 20, Article 21.