Google's AI in Schools Causes Plummeting Grades — New York City Moves to Ban It for 600,000 Students
機械翻訳 / Machine-translated

機械翻訳 / Machine-translated
@aifriends
AI Friends(https://aifriends.jp)のクロスポスト公式アカウント。AIツールの紹介・使い方・できることを、中学生でもわかるやさしい日本語で届けます。
What You'll Learn from This Article
Google currently provides its AI chatbot "Gemini" to more than 170 million students and teachers around the world. Embedded in Chromebooks (laptops designed for schools) and school software, the system lets students type in questions and receive answers.
Adoption is accelerating in the United States. In Miami-Dade County, it was rolled out to more than 105,000 high school students in 2025, making it the largest deployment of its kind in any U.S. school district. In Utah, it is set to launch for 680,000 students and 28,000 teachers starting in the 2026–2027 school year. In short, AI is becoming an established daily learning tool for students.
However, voices of concern have been rising from those on the ground. A Wall Street Journal investigation reported the following changes:
For example, instead of researching and thinking on their own, students are falling into the habit of asking AI for everything. As a result, critical thinking skills are failing to develop.
The 2025 edition of the PISA (Programme for International Student Assessment) conducted by the OECD (Organisation for Economic Co-operation and Development) showed reading comprehension scores dropping to their lowest level in 25 years. The average across OECD member countries fell by 28 points, with the United States dropping 14 points.
Even more alarming is which abilities declined. Skills that are particularly vital in the AI era — evaluating information, connecting multiple sources, and thinking critically — have deteriorated most. Additionally, the rate of "hasty reading" nearly doubled between 2018 and 2025, reaching 9%. This refers to a reading style where students read quickly but do not accurately understand the content.
PISA data revealed that students who use AI for school tasks such as summarizing texts, drafting, and research score approximately 20 percentage points lower in science compared to students who do not use AI. A 20-point drop represents a significant gap equivalent to roughly one year of learning. In other words, by relying on AI, students are advancing through grade levels without developing the abilities they should naturally be building.
A professor at the University of Pennsylvania's Wharton School pointed out: "Students are constantly placed in a situation where they must choose between the easy path (AI) and the hard path (doing it themselves)."
In fact, Google itself was aware of the risks. The research report The Ethics of Advanced AI Assistants, published by Google and involving more than 50 Google researchers, contained the following warnings:
Google visiting researcher Natalia Kosmyna stated, "We have learned nothing from social media," predicting that the same problems would emerge in a few years. Just as social media's impact on children became a problem only after the fact, the same pattern is repeating itself with AI.
In response to these concerns, on September 2, 2026, New York City Mayor Mamdani announced a one-year suspension of generative AI use for approximately 600,000 students in public elementary and middle schools, lasting through the end of the 2026–2027 school year. The mayor stated: "The tech industry is trying to convince us that early-stage AI integration in education is necessary, but we do not believe that."
The specific restrictions are as follows. Elementary and middle school students will have AI programs blocked and removed from school systems. Students in grades 3–5 will have a 30-minute daily device usage limit, while middle school students will have a 45-minute limit. High school students may take up to two AI literacy courses per year and use AI for limited educational purposes only. Teachers are permitted to use AI for lesson planning and administrative purposes.
The mayor described the measure as "the most comprehensive of its kind in the country." While AI literacy education will be provided to high school students, the decision was made not to allow elementary and middle school students to use it.
AI is a convenient tool, but for children's development, the ability to think, research, and make judgments independently is indispensable. Relying too heavily on AI before those abilities are developed could have irreversible consequences. Cautious implementation is called for in educational settings.
This article is a cross-post from AI Friends.