Abstract
Picture the overworked teacher: grading essays through the weekend, generating substitute plans to attend professional trainings, and crafting creative assignments to engage a classroom of students with diverse, often conflicting needs. Now consider the underserved student: sitting silently in the back of the room, too bored to take part and too shy to ask for help. Increasingly, this is the state of contemporary education.
Teacher shortages are nearing a crisis. In the United States alone, nearly 5,000 teacher vacancies are projected by 2025. 1 Other countries, including England and Germany, face similar challenges. Educators are leaving the profession due to increasing workloads, difficult conditions, and low pay. On average, teachers work 50-hour weeks, with less than half of that time spent directly interacting with students. 2 Today’s educators are overworked and underpaid, and children are the ones who suffer; luckily, recent developments in artificial intelligence (AI) offer promising solutions to alleviate some of this burden.
In 2022, ChatGPT was first released to the general public. 3 Other AI programs, such as Siri, the virtual assistant on Apple devices, had already been mainstream for years, but ChatGPT marked a major milestone: it popularized the use of generative AI, technology capable of generating content based on deep learning models. 4 Just a few short years ago, AI technology was not able to produce the high-quality images, videos, and text it can produce today. Since its release, ChatGPT and similar generative AI technologies have become not only part of the public consciousness but also key elements in discussions about the future of various industries, including education.
AI’s Role in Enriching Education
AI has already entered the educational sphere, with private sector companies developing and deploying technologies in schools and universities globally, creating a market valued at around $6 billion. 5 From helping teachers generate lesson plans to assisting students with complex concepts, AI is poised to have an increasing presence in the classroom.
One area where AI can shine is in personalization. Educators can now use this technology to tailor educational experiences to meet students’ diverse needs, adapting lessons to wherever a student may be in their learning journeys. For example, AI systems and tools already enable instructional sequences to adjust to student needs, offering feedback and hints during mathematics problem-solving or foreign language learning. 2 In addition, the anonymity of technology, paired with AI’s ability to conduct human-like conversations, may encourage students to ask questions and explore concepts without fear of judgment. This is especially beneficial for neurodiverse learners and students with disabilities.
For students who need more time to absorb concepts or who are not able to attend school physically due to socioeconomic or geographical circumstances, an adaptive chatbot tutor could be designed to explain difficult concepts outside of classroom hours, allowing students to learn at their own pace. One example of such AI-enabled tutoring technology is the intelligent tutoring system (ITS), which is modeled after how human experts solve mathematical problems. 6 Through ITS, students can receive feedback on specific steps in a problem’s solution process, rather than just an alert for a right or wrong answer. While still in its early stages and primarily tested in university settings, the expansion of programs such as ITS could provide individualized feedback to students as they work through challenging problems on their own time.
AI models can also complement classroom simulation tools, such as virtual reality (VR) headsets. AI-enhanced virtual reality creates interactive simulations, transforming VR into a pedagogical tool. For example, in higher education, VR allows medical students to simulate surgeries and procedures and study human anatomy in detail, reducing reliance on cadavers. In primary and secondary education, AI-powered VR can help students build confidence in public speaking, simulate different classroom settings for teachers in training, and bring lesson plans to life in new, engaging ways.
Beyond this, generative AI technology can reduce teachers’ mental loads by assisting with tasks like managing reminders, record-keeping, and classroom technology, freeing up time for teachers to have more meaningful interactions with students. AI can draft lesson plans, develop quizzes, and generate creative ideas for engaging classroom activities. AI algorithms are also capable of identifying patterns, alerting teachers when a student’s participation or performance changes, and signaling when the teacher should step in. For students with specific behavioral or learning needs, future AI programs could provide resources to parents for supporting their children’s education (like voice assistants that allow students to present their project using American Sign Language, which could then be audibly voiced for the class), 2 or even translate reports from high-stakes meetings for non-English-speaking parents. 7
Ethical AI Must be Human First
While AI presents an opportunity to transform education by enriching learning experiences and supporting educators, its integration must be carefully managed to address ethical concerns. The effective application of AI in education raises important questions, especially as it involves vulnerable populations. It is important to consider the social and ethical implications of bringing this technology into schools, the impact on teachers’ roles and classroom oversight, and how educational equity can be maintained without disadvantaging underrepresented student populations.
To effectively serve education, AI must operate with a human-first mindset. While it offers powerful tools to complement teaching, AI can never replace the vital connection between teachers and students. Educators bring cultural understanding, empathy, and situational judgment that algorithms cannot replicate. For instance, while an AI system may flag a student for falling behind academically, only a teacher can identify whether the underlying cause is an external challenge, such as family issues or undiagnosed learning difficulties. Similarly, teachers’ personal experiences and professional expertise provide essential context, allowing them to navigate the unique challenges of classrooms in urban, rural, or culturally diverse settings.
AI in education will require human oversight and ethical frameworks to preserve the humanity and care necessary for quality education in the digital age. Adopting technology that evolves faster than we can implement regulatory frameworks can pose significant risks. In education, incorporating AI into classrooms without fully considering how it may affect inclusivity and equity could have troubling negative effects, deepening the divide between those with access to technology and those without.
To ensure AI’s successful integration into the classroom, teachers’ needs and insights must be prioritized. Teachers bring a more comprehensive understanding of each learner than any educational technology can provide. For AI to succeed in easing teachers’ workloads, it must be developed with their input and supported by training. In fall 2023, 18% of K-12 teachers reported using AI for teaching—primarily virtual learning platforms, adaptive learning systems, and chatbots—and another 15% had tried AI at least once. 8 As these numbers grow, districts and companies must support teachers as they learn to incorporate these systems into their routines.
AI systems will need to provide information and recommendations without making critical decisions. Teachers should control how AI tools are used, whether it’s sending reminders to students or explaining key concepts. Educators will also need transparency about how AI systems make recommendations, the data they use, and how that data is stored. This is crucial because any AI tool in the classroom will have access to both teacher and student information. Concerns about surveillance may prevent teachers and students from wanting to use the technology—will districts be able to monitor their teachers’ activities and use that data to inform staffing decisions? Will students’ data be adequately protected from third parties?
Another concern is that students might use AI to cheat on assignments and exams, although data suggests that generative AI has not radically changed cheating behaviors in high schools. 9 However, AI could potentially undermine critical thinking skills and social connections, as students might rely on AI chatbots instead of working through difficult concepts by engaging with their peers and teachers. Mitigating these risks requires that teachers receive adequate training in the strengths and pitfalls of AI. Teachers need to understand when and how to use AI technologies to their benefit and how to carefully assess AI with an eye for quality control. Students should also receive training in how AI technology works, with clear instruction about when its use is appropriate. 8
The successful future of AI in education hinges on collaboration among policymakers, educators, students, parents, and developers. Together, they can create tools tailored to educational needs, reduce teacher workloads, and uphold rigorous privacy standards. Policymakers play a critical role in promoting fair access to AI technologies and setting safeguards to protect user data. When implemented thoughtfully, AI has the potential to empower both students and teachers, fostering an innovative educational environment.