These days, AI seems to have seeped into every corner of society. From advertising campaigns that push a company’s new ‘AI powered’ product to not-so-far-fetched Internet memes about humans losing their jobs to the new technology, we can confidently say we’re in a new era of advancement—and educational institutions are not spared that change. Although the technology is still in its somewhat nascent days, its far-reaching effects should prompt school administrators across the world to examine their AI policies, and, if they aren’t generally favorable towards the new technology, make them more so. Preparing students for an AI future necessitates a shift in thinking: from viewing AI as a threat to welcoming its potential as a transformative educational tool.

It’s no secret that students use AI for schoolwork on a regular basis. In addition to probably its best-known use, writing the good old English essay, newer versions can do more sophisticated tasks. Upload a photo of a math problem, and an AI model can break it down, step by step. Ask AI to make a slideshow about the Italian Renaissance, and it will. A few well-worded prompts can replace hours of textbook reading, so why are some people so hesitant to adopt the technology?

In defense of AI detractors, there is something to be said for pushing through difficult material without the easy cushion of AI to assist you. If reading a textbook description of Hemingway’s works doesn’t expose you to the acute simplicity of his phrases, then reading an AI summary of the textbook definitely doesn’t teach you much about his craft. Extremely distilled information, the kind AI tends to produce, is easy to consume but takes the consumer further and further from the primary source they’re trying to learn from, and knowledge is inevitably lost in the process.

But that downside is something that needs to be accepted in the quest to turn out better, more capable students. Schools must move beyond a stance of fearfulness when it comes to AI. The knee-jerk reaction to prohibit AI is not only unrealistic but also unfair to the students who will inevitably have to confront and utilize AI in their work and personal lives. It would be a very ill-informed assumption to think that AI is at its peak, that the technology won’t get any smarter, faster, and more applicable. By the time current and future ETHS students hit the workforce, AI will have completely reshaped almost every industry. If new job-seekers are coming from environments where AI has been willfully ignored, then they’ll immediately be at a disadvantage compared to those who embraced the new technology and the opportunities it created.

Even aside from marketability post-graduation, AI can and should be used in the classroom to enhance education. AI can offer individual support to students in different stages of learning, with personalized feedback and resources. AI writing software, for example, can guide students through the complexities of storytelling, posing questions that require deeper reflection and accuracy of their writing. This kind of individualized attention can be particularly helpful in a large and diverse school like ETHS, where students have varying needs and levels of access. By freeing students and teachers from mundane tasks by automating them, AI can enable both to focus on higher-order thinking tasks that facilitate deeper engagement with the content, making the learning more focused.

Also, AI can address issues of equity in education. While there will always be barriers to education in some form or another, AI, with its universally accessible format and increasing competency, can lower that barrier.

“I think [AI] has the potential to equalize education. It’s got a lot of potential to help us be more efficient, right? If a kid who doesn’t have access to a personal tutor, maybe AI can be that kid’s tutor, and it’ll be cheap,” said Superintendent Marcus Campbell. “I think it’s got the potential to level the playing field in a lot of ways.”

So, how should schools implement AI in their classroom? The bad news is that there is no set, magic curriculum that will turn out AI-powered yet humanely thoughtful students. The good news is that there are a few essential ideas that can be implemented in school curricula to adapt to the AI world.

Just like news literacy is taught in many school districts, AI literacy can and should be taught as well. Knowing how to analyze the word choices and associated biases of news articles makes for much better news consumers, and the same principle applies to AI. Students should learn how AI models are created, maintained, and improved. Students should learn how to properly write prompts for AI, and, at a deeper level, how to work with even more complex models than the relatively accessible ChatGPT (and related models). Students should learn how the biases of creators are often ingrained in AI, and how to fact check the responses they get from using it. AI literacy is something that is easily taught with the right curriculum, and will make students more capable AI consumers.

The other key concept that goes hand in hand with AI implementation in the classroom is the idea of using the technology to supplement—not replace—human instruction. When a student is using AI, there are always bits of information they need to know from the outside world to make their AI experience as productive as possible. If a student never learned from a teacher what makes a good English essay, how are they supposed to judge the output of an AI engine? There are also key moments in the classroom where critical thinking needs to happen completely AI-free. If we continue with the English example, some of the best critical thinking a student does can happen during in-class essays, where they have an hour to analyze a text and write a cohesive, but brief, paper about it. In that situation, it’s obvious that AI could churn out a good enough paper in five minutes, but that misses the real point of forcing students to push themselves mentally while writing the essay. For both those reasons, it’s evident that AI has a specific time and place in the classroom, and in the future educators need to find that balance between AI and human instruction.

Fortunately, ETHS is already taking some steps in that direction. “We’re training teachers to be more embracing of AI, what it can do to assist in instruction, as opposed to what some schools have done and just banned AI, which I don’t think is probably reasonable or practical at this point,” said Campbell.

AI is here for the long haul, and by teaching students how to correctly use it, we can make it seem less like an antagonistic world-ender and more as an extremely helpful, adaptive tool. At ETHS, and at centers of education across the world, we need to position ourselves as pro-AI.

“I don’t think AI is a threat to human intelligence,” said Campbell. “I think it’s a form of human ingenuity.”