As schools across the United States increasingly incorporate artificial intelligence (AI) tools into classrooms, concerns are mounting over their unregulated use and the potential impacts on student learning. Education Secretary Linda McMahon recently defended AI integration during an appearance on CNN’s “State of the Union,” citing a private school model in Austin, Texas, where students reportedly spend two hours daily on personalized learning through AI-driven platforms. However, experts and research suggest a more cautious approach may be necessary.
McMahon highlighted Alpha School, a private institution co-founded by an AI entrepreneur, as an example of AI’s potential in education. Students reportedly work independently on computer-based lessons while teachers monitor progress, creating a near one-on-one tutoring environment, according to McMahon. She described the approach as effective in that setting, noting it had been in place for several years. Yet, Alpha School’s results have not undergone independent verification, and the private school’s selective admissions practices complicate broader applicability.
Critics point to mixed evidence regarding AI’s impact on academic outcomes. Sal Khan, founder of the educational platform Khan Academy whose AI chatbot Khanmigo has been introduced in public schools, acknowledged the tool had a limited effect on many students. Reports have also emerged of families withdrawing children from Alpha School locations due to dissatisfaction with both academic and social experiences.
A significant study released in June by the Center for Economic Policy Research analyzed over 26,000 students in China and found AI use correlated with improved homework performance but a decline of roughly 20% in monthly exam scores over six months. Entrance exam results also dropped between 18% and 24%, suggesting some students may rely on AI for assignments at the expense of deeper learning.
The debate over AI’s role in education unfolds amid concerns about Big Tech companies’ influence on policy and the classroom. Technology firms like Google, Microsoft, Apple, and OpenAI are active in promoting AI adoption through funding research, shaping curricula, providing instructional materials, and participating in educational committees. Observers warn this close involvement may skew priorities toward corporate interests rather than pedagogical effectiveness.
Experts advocate for a more comprehensive framework to evaluate educational technologies. Natalia Kucirkova, a professor specializing in early childhood development at the University of Stavanger in Norway, proposes assessing ed tech tools across five dimensions: efficacy (impact on learning outcomes), effectiveness (real-world classroom use), ethics (data privacy and security), equity (accessibility for diverse student populations), and environmental sustainability. She emphasizes that technologies must first demonstrate measurable benefits and practical utility before other considerations.
Kucirkova also highlights the challenges of implementing such standards within a profit-driven educational technology market. She supports increased government investment to develop educational tools as public goods, drawing parallels to the collaborative creation of the educational television program “Sesame Street,” which yielded documented learning benefits globally.
In contrast to the U.S., Norway has taken a more conservative stance on AI in schools, restricting use for children under 13, promoting supervised adoption for those 14 to 16, and encouraging responsible AI literacy for students aged 17 to 19. While the Norwegian model reflects a preference for prudence and public trust, experts acknowledge the U.S. faces greater diversity and systemic challenges in regulating technology.
The discussion on AI in education continues to evolve rapidly. Advocates urge policymakers to establish rigorous evaluation standards and oversight mechanisms to ensure that emerging technologies serve the interests of students and educators rather than primarily advancing corporate agendas. Without timely and thoughtful regulation, there is concern that educational environments may become testing grounds for unproven AI applications at the expense of learning quality.