Almost three years into the ChatGPT frenzy, and we’ve heard it all about AI. Ever since ChatGPT was released, and kids started writing essays with AI, the hype and horrors of chatbots have taken over conversations about teaching and learning.

Will teaching students AI skills really help them in the future, or is it just an arresting tool right now? Are there other skills that would better prepare them for the future workforce?

From replacing teachers, to AI study buddies, to students advocating for ethical AI, EdSurge has explored the challenges and potential for AI in education for the past few years. In this research project, we wanted to dig deeper and ask critical questions about AI usage in schools. With all we’ve learned about AI over the past few years, we wanted to talk to some teachers and even had some of them test lessons on AI with their students to get their thoughts.

In Fall 2024, EdSurge Research partnered with Rethinking Schools, an education nonprofit and publisher focused on social justice and education, to conduct a participatory research project about how artificial intelligence is changing teaching and learning. The purpose of this project is to share resources with teachers and to add their nuanced perspectives to the very loud conversations about AI in teaching and learning.

For our research project, we gathered a small group of 17 teachers and created a forum to encourage critical conversations about how artificial intelligence is impacting their classrooms to better understand the challenges and opportunities of AI tools in teaching and learning.

Our general research questions were exploratory in nature to facilitate a space for teachers to 1) talk about their opinions on AI, 2) test generative AI for teaching and learning, and 3) critically assess the possibilities and pitfalls of AI integration.

Some of the questions we considered in our exploration were:

• What is the role of AI in supporting student learning?
• How are teachers using AI to foster student engagement and motivation?
• How do we leverage the power of AI to better support teachers?
• Where is AI already having a positive impact on schools and making teachers’ jobs more efficient?

Based on the perspectives shared by teachers who participated in the small group conversation and our research projects, we believe the insights they shared can help us contextualize the hype and horrors of the AI industrial revolution.

What AI Research Says

The Concern: Are adults’ concerns about students substituting AI for critical thinking legitimate?

In October 2024, Pew Research surveyed teachers about their views on AI’s impact on students. A third of teachers reported that they believe AI is equally or more harmful than beneficial for students, and a third of the teachers weren’t quite sure yet. Meanwhile, Pew Research’s recent student survey from January 2025 showed that most 13- to 17-year-olds still haven’t utilized ChatGPT for school; however, ChatGPT usage for schoolwork increased from 13 percent in 2023 to 26 percent in 2024.

A school year later, in another teacher survey from the Walton Family Foundation, teachers report an almost equal split between feeling favorable, indifferent and opposed to AI. Approximately 48 percent of teachers who use AI at least weekly think that AI tools will help students with problem solving, critical thinking and engagement.

However, industry workplace studies have shown mixed results about the productivity, efficiency and effort that using generative AI takes or gives back to employees. While anecdotal users’ and marketing campaigns tout generative AI’s time-saving potential, the mixed results of recent studies have challenged that assumption. In a small sample randomized control trial, seasoned tech employees spent more time on coding tasks when using generative AI for assistance. A larger study of about 5,000 computer programmers at Microsoft, Accenture and an anonymous Fortune 100 tech company showed mixed results. Although the average participant saw a 26 percent time boost when using an AI coding assistant tool, the researchers note that there was a lot of variability in this number across roles and experience levels. Both studies have limitations and focus on today’s tech workforce, but these findings could have implications for teaching and learning.

Some experts say AI should be integrated into classrooms to prepare students for the future. Others say AI skills aren’t integral to students’ future success. A 2023 report from the International Computer and Information Literacy Survey tells this story: While essential computer and information literacy for eighth grade students in the U.S. is about average compared to other nations, American computational thinking scores lag the international average by 22 points.

Can AI and Critical Thinking Coexist in Classrooms?

We don’t know yet.

As AI evolves and schools use it more, we’ll have more opportunities to assess how frequent generative AI usage affects learning and cognition. Beneath the surface of the cheating concerns, one of the greatest issues with scaling AI in schools is risking declining cognitive development and critical thinking skills. Several emerging studies show that adolescent AI usage can interrupt key components of cognitive development, like active recall, problem solving and creativity.

Our Research Project

We utilized the participatory action research and case study methods for this research project. We wanted to offer teachers foundational tools and resources about integrating AI into their classrooms while capturing their voices in a conversation they been included in before. That’s why the participatory action research was effective for this exploration — teachers could connect, share hopes and concerns and bounce ideas off one another.

A French teacher, elementary teachers, math teachers, chemistry teachers, computer science and engineering teachers — we talked to 17 teachers from half a dozen countries and every region of the U.S. who teach grades three to 12. In our forthcoming stories, we’ll share quotes from the discussions during our virtual learning circles. Then, we’ll share more details from our case studies about how some of the teachers’ lessons went.

Looking Forward to Teaching Tech

In November 2022 ChatGPT was released, and a few months later, students were using it. Teachers were discovering that students were using it, then companies discovered that people were using it. And suddenly, AI-integrated tools were advertised all around us. While this is the most recent massive mainstream buzzword takeover, I predict it won’t be the last.

The education ecosystem is loud. There are experts outside the classroom – consultants, speakers, nonprofits pushing in, professional development providers and researchers like me. The narrative around what’s best for education is sensitive to the technological industrial revolution. Experts outside the classroom want schools to quickly adapt to change, while the people inside the classroom are closing the door to the noise.

Like education experts warn about “teaching to the test,” are we teaching with tech, or teaching to the tech? Should we encourage educators to seismically change curriculum for every new tool that’s created? Some change is good — great, even. But what if generative AI chatbots, in their current iteration, aren’t ready to be centered? Scaling generative AI in the U.S. has grave consequences for pedagogy, engagement, cognitive development and the environment.

In our upcoming EdSurge Research series, Teaching Tech, we’ll explore these questions and share more about what we learned from our research project.

For a selected list of peer-reviewed articles, research and data studies referenced in this story, click here.