How to use AI as a learning partner, not a replacement

By: Joey Morris

Published: January 21, 2026

When you hear the words "AI" and "education" together, you may envision learners generating entire papers with ChatGPT and outsourcing their thinking processes. While these shortcuts exist (and are being used), the reality of AI's use in education is much more nuanced.

Recent research from College Board shows that the percentage of high school students who use generative AI tools for their schoolwork increased from 79% to 84% between January and May 2025. Furthermore, nearly 6 in 10 parents agree that students should use generative AI. With the growing emphasis on AI skills in the workplace, it is easy to understand why.

However, a study published in the journal Computers and Education: Artificial Intelligence found that the majority of students were more likely to perceive the use of AI as acceptable for generating ideas for a paper, project, or assignment than for completing all or part of the work itself.

As these examples illustrate, we now live in a time when most learners have access to and the ability to use AI to augment their learning. They are already forming their own opinions about what constitutes acceptable and unacceptable use cases. The critical challenge that educators and learners now face is finding and encouraging the appropriate balance between effort and assistance.

To gain a deeper understanding of what is at stake, it's essential to identify the risks and benefits of AI-assisted learning.

What's at risk? Your learning on large language models

While generative AI offers powerful educational support, it also introduces the temptation to outsource your thinking. Adam Hill, a principal learning strategist with edX, is thinking about the future of education and how to help learners avoid these pitfalls through the newly updated "How to Learn Online" course on edX.

Cognitive offloading

One of the primary risks associated with relying heavily on AI is cognitive offloading. Hill describes cognitive offloading as "the process of reducing one's mental load by taking actions to change the demand of a task. It's finding ways to avoid holding information in your mind to free your brain to do other things."

Cognitive offloading is not a new phenomenon, though. "We have always done this throughout human history — writing things down is a form of cognitive offloading. One of the examples that often gets referred to is the use of the calculator." So what makes AI so different from the technologies and tools we use every day?

"Cognitive offloading is the process of reducing one's mental load by taking actions to change the demand of a task. It's finding ways to avoid holding information in your mind to free your brain to do other things."

— Adam Hill, principal learning strategist at edX

Hill explains that AI is a significantly more transformative tool than a notebook or a calculator. "You may use a calculator to offload a smaller mental task so that you could think more about higher-order thinking skills around mathematics, but AI allows you to potentially offload almost all of your thinking to another tool." This can lead to stunted thinking skills and codependence.

In a preliminary study published by MIT, researchers found that the use of large language models (LLMs) has the potential to reduce brain engagement and memory recall. AI is still a new technology, and more research will be required. However, "the potential long-term impacts of extensive cognitive offloading with AI are largely unknown," Hill cautions.

Desirable difficulty

Closely tied to cognitive offloading is the concept of desirable difficulty. According to Hill, "desirable difficulty is the idea that good learning happens when learning requires effort."

"It is the result of offloading too much cognitive work that learners miss out on this mental effort that is essential for strengthening their own thinking skills, developing metacognition, and building genuine long-term understanding."

"Desirable difficulty is the idea that good learning happens when learning requires effort."

— Adam Hill, principal learning strategist at edX

The ease and speed of generative AI tools like ChatGPT pose a tempting alternative to a more effortful learning process. "The concern here is that learners will too quickly resort to AI tools rather than experiencing that productive struggle — attempting tasks, making mistakes, and ultimately developing the cognitive skills and mental frameworks necessary for true understanding."

Recognizing these risks empowers learners and educators alike to consciously engage with AI tools in ways that enhance learning, rather than unintentionally replacing critical thinking and deeper cognitive engagement.

Practical strategies to learn effectively with AI

Like any new technology, AI comes with risks, but it also presents tremendous upsides. Fortunately, there are effective strategies that you can use to engage with AI productively without compromising your learning.

1. Leverage AI for learning personalization

One of the greatest benefits of using AI is its ability to improve engagement through personalization. AI can tailor learning materials to your personal interests, skill level, and preferred learning methods. Hill provides the following example.

"Say you're having difficulty understanding something. You can ask AI to explain it to you as if you're a fifth grader. If you're finding a topic kind of dry, you can ask AI to explain it through the lens of something you're more interested in." This type of learning can allow you to engage more deeply and effectively with the material.

From an instructor's perspective, this level of personalization can also be effective, but it is often not feasible at scale. With AI, learners can benefit from personalized learning on their own terms.

2. Engage AI for active testing and retrieval practice

Reflect on how you used to study for an upcoming test. You may have reviewed your notes or highlighted material in the textbook. These are examples of passive learning, which are not as effective as more learner-centered, active learning practices. Active learning involves a higher level of engagement, such as quizzing yourself with flashcards or working through a problem independently.

Hill recommends using AI to better facilitate active learning. For example, you can ask AI to quiz you on certain lessons, create custom flashcards based on a recent chapter, or even identify and focus on potential knowledge gaps in your learning. In addition to being more effective, you're also likely to find this type of learning more engaging.

3. Employ AI for brainstorming and idea generation

AI tools can be used to expand on an original idea and offer suggestions for ways to move forward. "There is the idea that creative people don't have better ideas, they just have more of them," says Hill. "So if you have an idea for a paper and you ask AI to give you 50 ideas on how to write this paper, that can be very helpful, as long as you still select and synthesize among those choices."

4. Use AI to support feedback and revision

Hill describes AI as akin to having accessible "office hours" at any time. "When I was in college, I tried — as intimidating as it was — to go to my professors to get feedback." While AI isn't a replacement for office hours, it can be a helpful tool for reviewing your work or providing suggestions. These suggestions shouldn't be accepted blindly, though. "The thing to remember with feedback is that you still have to decide whether you're going to incorporate it or not," says Hill.

5. Prioritize understanding over task completion

AI makes it tempting to rush through assignments, completing tasks quickly rather than deeply engaging with the material. However, speed doesn't guarantee effectiveness. Hill emphasizes adopting a "process-oriented approach to assignments rather than a product-oriented approach." In other words, try not to view your assignment merely as the final product; you have the most to gain from the learning process itself.

Before relying on AI, always ask yourself: Is my goal to complete this task as quickly as possible, or to genuinely understand it? Prioritizing true understanding ensures that you gain meaningful knowledge and retain information, rather than merely completing another assignment.

6. Develop critical evaluation skills to assess AI outputs

As AI becomes more sophisticated, it's increasingly important to develop strong evaluation skills. Hill emphasizes the importance of critically assessing AI outputs for bias, accuracy, and quality.

For instance, tools like Deep Research can rapidly generate extensive papers — sometimes up to 40 pages of content. Hill points out that even experts may struggle to thoroughly verify such dense AI-generated content. He suggests approaching these outputs thoughtfully: You don't need to use every detail, but rather selectively identify helpful information, relevant citations, or valuable insights. The key is to remain critically engaged and discerning when leveraging AI for research and learning.

7. Utilize AI for administrative and organizational tasks

Lastly, consider this practical suggestion: Use AI for what it currently does best — handling the more mundane or repetitive tasks. Hill provides the example of creating a works cited page, a task that AI can streamline effortlessly. Other examples include formatting essay outlines, organizing lecture notes, or summarizing lengthy readings.

Delegating these tasks to AI can save you valuable time and mental energy, letting you direct your focus toward more meaningful aspects of learning. However, it's crucial to regularly pause and ask yourself: Am I using AI in a way that allows me to concentrate on the core aspects of my learning?

Looking ahead to learning with AI

Whether you plan to rely heavily on AI in your learning or not at all, it's important to be intentional and understand how this technology is shaping education.

Hill is excited about what the future holds. "It's critical that students and educators stay tapped into this moment and think about the innovative ways that they can both use AI for teaching and learning. We might see new ways of learning emerge in this moment. It's a really exciting time to be an educator."