Teachers Believe That AI Is Here to Stay in Education. How It Should Be Taught Is Debatable.

One of the perks of Angie Adams’ job at Samsung is that every year, she gets to witness how some of the country’s most talented emerging scientists are tackling difficult problems in creative ways.

They’re working on AI tools that can recognize the signs of oncoming panic attacks for kids on the autism spectrum in one case, and figuring out how drones can be used effectively to fight wildfires in another.

What’s remarkable about these innovations is that most of their creators aren’t old enough to get their driver’s licences yet. They’re part of Samsung’s Solve for Tomorrow tech competition for public middle and high school students, and winning means big prize money for their schools to purchase more tech tools.

While every year’s crop of finalists is impressive, Adams says the program organizers noticed something different about this year’s cohort of brainy student engineers.

Among the top 50 teams, 42 percent used artificial intelligence to power their inventions. That’s up from just 6 percent in 2024.

That’s a good thing in Adams’ view, as she’s more than a little confident that today’s K-12 students will be using AI in some fashion when they eventually join the workforce.

“I rarely say 100 percent to anything, but I think the answer is 100 percent,” Adams, senior manager of Corporate Citizenship at Samsung, says of the proportion of students who will use AI at work in the future. “We really believe that that is something that starts in the classroom, so we want to make sure we’re doing our part to ensure students have the skills to understand, use, and create with this emerging technology.”

At the classroom level, teachers are on the same page.

In a survey of more than 1,000 public school teachers — done through Samsung’s partners at DonorsChoose — a whopping 96 percent said AI “will become an intrinsic part of education within the next decade.” Just as many said they currently lack the resources to integrate artificial intelligence into their curriculum.

Educators were generally optimistic about the use of AI in the classroom, with more than half saying they already use it and another 33 percent saying they are exploring how they can integrate it. The most common uses of AI teachers reported were to personalize student learning, interactive learning tools and getting insights on student performance with data analytics.

Their most common concerns about AI were plagiarism, a lack of teacher training on AI education tools, its potential to spread misinformation and a reduction of student interaction during class.

Few teachers were worried about losing their jobs to AI — just 5 percent were concerned.

While Adams predicts that students will use AI in their careers and as teachers experiment with its use in their classrooms, more school districts are moving to formalize AI in their curriculum.

Zarek Drozda, director of the nonprofit Data Science for Everyone, says his organization has seen interest in offering AI and data science coursework increase among school districts, with the number of states launching data initiatives increasing from one to 29 over the past four years. Data science concepts form the building blocks of artificial intelligence, including popular large language models like ChatGPT.

“We’ve seen fast growth of state pilots, professional development programs that are state supported, standards revisions,” Drozda says, “[and] the curriculum market for data science and data literacy and AI literacy is growing quite quickly. We are seeing a lot of interest from school leaders.”

He says the appeal of data science is that it provides a concrete onramp for students to learn about artificial intelligence.

“I think data science also provides a very compelling framework for students to evaluate AI tools with some skepticism and understand the use cases deeply,” Drozda says. “ChatGPT is trained on text data. It’s going to be really good for writing, not so good for math, as one example.”

Drozda says schools don’t necessarily need to rush into building AI classes or programs. They can start with mastering spreadsheets, coding languages like Python or teaching students to use AI chatbots.

“No school leader should think they have to do everything all at once. It is completely OK to take small, entry-level steps to begin to prepare everyone for the broader technology landscape,” Drozda offers. “I think the way that the data science and data literacy movement in particular is approaching this is through bite-sized modules. Try two weeks on a concept in mathematics, try this data set to cover the existing unit you already have on ecosystems in biology, teach the booms and busts of economics through data from the Federal Reserve.”

Districts aren’t only thinking about AI as part of teaching — they’re exploring how it can help with a wide swath of jobs.

Pete Just is the generative AI project director for the Consortium for School Networking, a professional association for K-12 edtech leaders. The organization has created a suite of guidelines to help school districts think through their artificial intelligence usage and policies.

Just says the life cycle of teachers’ feelings about AI started at confusion, then fear about it threatening their jobs, followed by worries about students cheating but also a desire to see how the technology can be useful with lesson planning and other administrative tasks.

“Wherever something new happens, that’s the initial attitude,” he says of teachers’ early skepticism, “but generative AI has made an impact that you cannot deny.”

Generative AI has the potential to help districts operate more efficiently, from managing spreadsheets to bus schedules.

“Now we can do things that help families and students better than in the past,” Just says. “When we get to student chronic absenteeism, connecting that to a database with student absences and making connections with parents to give them a day-to-day update on [whether] their student was here or not, that communication window is so much shorter because you don’t have to make the phone call.”

But districts also have to contend with something that the “move fast and break things” ethos of the Silicon Valley culture behind AI developments do not: Schools’ legal obligations around protecting student data.

Beyond artificial intelligence’s potential to make administrative processes smoother, there are a host of logistical and ethical considerations that Just says districts must make when it comes to infusing AI into their curriculums.

The basis of any AI education would have to be rooted in critical thinking, he explains, how to ask good questions and evaluate the answers.

“You look at the results and say, ‘That doesn’t even match up with reality of what I know about this historical event,’” Just says. “Because sometimes it hallucinates, being able to say that makes sense — or makes no sense — becomes important.”

Ultimately, Just isn’t a fan of making artificial intelligence its own stand-alone class. Instead, he believes it’s something that needs to be integrated into every class so that students can learn how to apply it within each discipline.

“The easy thing to do is develop a class. You just need the school board to approve it, in a month and a half you can be done,” Just says. “What you’re really looking at is fundamentally changing the way you teach things, which no one wants to hear about because that’s really hard.”

That kind of AI integration is a level of complexity that he concedes will not be attractive to most districts. It’s a process that he says would take years of training teachers on integrating AI into the curriculum — about three to five years “if you’re working hard at it.”

“If you’re not working hard at it, you’re going to fall behind, and you’re not serving your students well,” Just says, “because in three to five years, every business is going to expect it. Even today, many businesses expect students coming out of high school to have the skill to be able to use these tools in the workplace.”