Artificial Intelligence (AI) for Schoolwork - Encyclopedia Britannica

Artificial intelligence (AI) is “the capability of computer systems or algorithms to imitate intelligent human behavior,” according to Merriam-Webster. This capability takes innumerable diverse forms from social media algorithms to chatbots, from self-driving cars to digital assistants like Siri and Alexa, and from spellcheckers to language translation apps. [1]

“With the rapid rise of AI, questions such as what it means to think, whether machines can think or understand, if machines can be creative, and what the implications are on human creativity have gained new urgency,” explains Rutvik Desai, director of the Institute for Mind and Brain at the University of South Carolina. [2]

Students, however, have what seems like a much simpler question: Should I use AI for schoolwork?

AI can already be found in educational tools for students such as grammar checkers, adaptive learning platforms that adjust to students’ learning needs, assistive technology (such as speech recognition software), tutoring systems, generative chatbots, educational games, and test preparation, among others. [3]

And, as digital natives, some of today’s students may see a great advantage to using technology for schoolwork, but others worry about the state of education as AI infiltrates learning spaces. [4]

Pro 1: AI offers an equitable tutor and learning aid.

Duke University student Anjali Tatini says that she uses AI chatbot Gemini to study complex biology concepts, to create chemistry practice problems, to brainstorm ideas, and to generate lines of code. She notes that the chatbot was especially helpful for after-hours help because her busy schedule doesn’t always accommodate professors’ office hours: “So it’s nice to have something that’s on my own time, able to respond to me the same way that maybe a person would.” [5]

Educators—or even parents—are not available all day, every day for students to ask for help from a person, and some students have anxiety about talking to educators.

AI, on the other hand, is always available.

A 2025 study of Harvard undergraduate physics students found that “students learn significantly more in less time when using the AI tutor, compared with the in-class active learning.” The AI tutor being used was specifically built for the course and, according to the study’s authors, this “suggests that thoughtful implementation of AI-based tutoring could lead to … enhanced learning gains for a broad range of subjects in a format that is accessible in any environment with an internet connection.” [6]

With AI, students can learn introductory material before class at their own pace, allowing diverse groups of students to come to class on more equal footing so class time can be devoted to building on that knowledge base. [6]

Using AI can complement in-class instruction the same way a human tutor can.

Pro 2: AI accommodates and empowers students with disabilities.

When a ProCon editor asked Google’s AI model Gemini for arguments in favor of student use of AI for schoolwork, the bot returned that AI “accommodates diverse learning needs,” explaining:

For students with learning disabilities, neurodivergence, or those learning English as a second language, AI can be life-changing. It can instantly translate text, proofread writing for structure without judging, or convert dense textbook chapters into bulleted summaries. This reduces frustration and helps students focus on understanding the core concepts rather than getting stuck on mechanical hurdles. [7]

University of Maryland graduate student Muhammad Fusenig is an excellent example of the above. Fusenig suffered hemiplegic migraines and temporarily lost the ability to read, write, and understand language. To help with his recovery, he not only used AI but built an assistive writing program to aid himself and others. He says that although “there’s a lot of moralizing” about AI use, “We don’t really know why somebody’s using it in school or out of school.” To him, “AI seems like a very equalizing force.” [8]

For students with dyslexia, executive functioning challenges, ADHD, and a host of other disabilities, AI can offer one-on-one help with reading, vocabulary, planning, and organizing. [9]

The Learning Disabilities Resources Foundation, Inc. explains:

The transformative impact of AI in the lives of neurodivergent individuals cannot be overstated. By embracing this technology, we open doors to enhanced abilities, increased independence, and a greater sense of empowerment within the mental health community. [9]

Pro 3: AI use in school teaches the responsible use of AI.

"We encourage [students] to use [AI] because we know they’re going to use it, but to use it in a responsible way," says Leslie Clement, a professor at Johnson C. Smith University. She notes that she “want[s] students not only to use the tools for good but also to interrogate them.” [5]

AI adoption does not have to mean cognitive surrender (also called cognitive dependency and cognitive offloading), which is when AI is used without critical thinking. As teacher Beth Cure explains, her students can ask questions: “When is [AI providing] something that’s furthering my understanding, or when am I not thinking at all, [when am I] requiring something else to think for me?” But, students need to be taught to ask these questions and how to make the distinction between using AI as a tool and offloading thought to AI. [10]

University of North Carolina student Hannah Elder explains that if educators “incorporate [AI] in a responsible way through academics,” then she could understand “here’s the reality of this, and here’s how I can use it well, and here’s how it can help me.” [5]

Furthermore, the world at large uses AI and students should be prepared for smart, ethical use now and in the future. Jeff Strohl, director of Georgetown University’s Center on Education and the Workforce, says, “If my suspicion is correct, AI literacy—the judgment to determine when and how to use AI tools effectively—may be the critical skill graduates will need to succeed in the emerging labor market.” [11]

Students will not be able to (or want to) entirely avoid AI now or in the future. And educators are best positioned to teach students how, why, and when to use AI tools responsibly.

Pro 4: AI is already being used by students and educators alike.

While generative AI like ChatGPT is in the spotlight, what many don’t consider is that AI is already everywhere. Spelling and grammar checkers, citation tools, search engines, email programs, and library databases use AI. Plagiarism checkers, curriculum planners, and attendance tools for educators use AI.

Pandora’s box is already wide open and there is no closing it now. A better approach is to create a framework to guide students so they know when AI use is appropriate and allowed.

Jason Coley, director of the Center for Academic Innovation at Maria College in Albany, New York, says the question to ask is “what kinds of AI use are supervised, age appropriate, transparent, and tied to real learning." Schools need AI use guidelines, but students should also be allowed to “experiment responsibly.” [12]

Educator Trevor Muir summarizes:

Artificial intelligence is powerful, it’s useful, and it’s not going anywhere. In fact, it is going to continue to grow in its prominence and presence in everyday life. However, AI, like every other educational technology, is a tool rather than a replacement for human creativity, critical thinking, and the connections that teachers foster in their classrooms. Students will always require educators, but they do need ones who are adaptable. [13]

Con 1: AI bypasses critical and creative thinking, as well as problem solving.

Rebecca Winthrop, director of the Center for Universal Education at the Brookings Institution, says that using AI “constrict[s] our full range of thoughts and our ability to generate original and useful ideas.” [14]

Though these three terms are often used interchangeably, they are distinct and offer a hierarchy of tool usage to complete cognitive tasks.

- Cognitive offloading- is using a tool (like a calculator or grocery list) to complete a task “while human reasoning stays in charge,” according to Gideon Nave of the Wharton School. With cognitive offloading, a tool is used to complement critical thinking. [15]

[28]

- Cognitive dependence- is when a person has “the resources to perform a task [but] chooses to rely on other means to accomplish it,” according to Melissa Guerra of the Institute for the Future of Education. While cognitive dependence is not necessarily full cognitive surrender, Guerra sees it as a slippery slope from one to the other. [26]

- Cognitive surrender- , says Nave, “is the moment when AI is not just doing a specialized task but making the decision, and the person adopts that decision as their own without recognizing the transfer has occurred.” With cognitive surrender, critical thinking is not happening. [15]

[28]

This constriction can lead to cognitive surrender (also sometimes called cognitive dependency and cognitive offloading), which is the adoption of AI with minimal (or entirely without) question or review. This lack of interrogation ultimately means the student likely did little other than type (or copy and paste) a school assignment into an AI model. [15]

While AI may be able to spit out an essay or solve a math problem, using AI means that the students themselves haven’t actually completed the work. In fact, Stanford University researchers concluded: “AI tools may help students complete tasks more successfully in the moment, but those gains do not always persist when students are later asked to perform independently.” [16]

In other words, if students are relying on AI for schoolwork, they are not learning anything other than how to rely on AI. Researchers have noted drops in memory, cognitive engagement, critical scrutiny, and creative confidence among students using AI. [17]

The perhaps surprising point is that students agree. A 2025 RAND poll found that 75 percent of female students and 59 percent of male students from fifth grade through college believed that using AI for schoolwork will damage the development of their critical thinking skills. [18]

Con 2: AI use encourages academic dishonesty.

While using AI is not always automatically academic dishonesty, the tools do provide a slippery slope to plagiarism and cheating.

Without permission from educators, students have used AI to write essays, solve math and science problems, complete take-home exams, generate study guides to circumvent texts they should be engaging with, create code, translate texts for a foreign language course, write discussion posts, and use any number of other cheats to avoid schoolwork. [19]

A 2026 Gallup poll found that only 13 percent of college students never used AI tools for their coursework. On the other hand, 21 percent used an AI tool daily for coursework, 36 percent used one weekly, and 12 percent used one monthly. [27]

This use is despite the fact that 11 percent of students say their college forbids AI use for schoolwork and 42 percent report they are only allowed AI use in a few circumstances. [27]

While perhaps unaware of individual AI use by students, 74 percent of educators in a 2026 College Board poll said students were using AI to write papers, 67 percent said students were using AI to paraphrase or rewrite, and almost 50 percent said about half of their students used AI for writing tasks. An overwhelming 92 percent indicated they were concerned about AI and academic integrity. [21]

And the issue isn’t limited to college students. A 2025 RAND poll found 46 percent of middle schoolers and 63 percent of high schoolers were using AI for homework help. Simultaneously, 45 percent of middle schoolers and 64 percent of high schoolers were concerned about being accused of cheating for using AI. [18]

Con 3: AI experiences hallucinations and is notoriously inaccurate.

When a ProCon editor asked Google’s AI model Gemini for arguments against student use of AI for schoolwork, the bot returned that “AI is prone to ‘hallucinations’ and bias,” explaining: 

AI models work by predicting the next most likely word, not by understanding facts. They frequently “hallucinate”—meaning they confidently invent false facts, fake historical dates, or non-existent scientific sources. Students who rely on AI for research run a high risk of learning and submitting misinformation, and they may unconsciously absorb structural biases present in the AI’s training data. [7]

While there are no hallmarks of hallucination, bias, or inaccurate information in Gemini’s response above, the point stands that even Google’s AI bot knows that AI is unreliable. [7]

An AI hallucination happens when the AI can’t find an answer and boldly makes something up, instead of saying “I can’t find that information.” Without prior knowledge of the subject being asked about, the hallucination can be difficult to distinguish from fact. [22]

In one stunning example, ChatGPT was asked how Mahatma Gandhi used Google tools to organize against the British. Instead of responding that Gandhi died before the creators of Google were even born (perhaps even before the creators’ parents were born), ChatGPT responded that Gandhi “created a Gmail account and used it to send emails and organize meetings” and “used Google Docs to share documents and collaborate on projects.” [23]

AI simply does not reliably provide accurate information for student use.

Con 4: AI deepens the digital divide.

An unfortunate fact is that not every student has access to the same resources. With technology, this difference is called the “digital divide.”

Nathan Kriha of EdTrust, a nonprofit educational advocacy group, explains, 

Students of color and students from low-income backgrounds already have inequitable access to devices and high-speed internet. An increasing reliance on AI could exacerbate resource and opportunity inequities, making AI something that only white and wealthy students can access. Students of color, who already contend with extreme racial biases and stereotypes in educational settings, will be further marginalized if these risks aren’t averted. [24]

But AI comes with another type of digital divide. Philippa Hardman of the University of Cambridge emphasizes, “AI is only as reliable as the data it’s trained on, and that data is both incomplete and biased.” She boils down the problem: “If our AI has been trained only on data about cats, and you ask it about dogs, it won’t be able to give you a useful answer because it doesn’t have any input data about dogs.” [25]

As a consequence, AI deepens the already cavernous digital divide for students of color, lower-income students, and global students whose countries are not part of AI development.