社交媒体的教训并非在于变革性技术应当被恐惧，而在于当儿童涉及其中时，防护措施必须先于伤害而设立——而非事后补救。
The lesson of social media is not that transformative technologies should be feared.
多年来，社交媒体平台在缺乏对儿童安全、透明度、适龄设计、家长知情权及独立监管等实质性保护措施的情况下野蛮生长。只有在危害证据不断积累后，政策制定者才开始认真应对。那时，数百万儿童已深受那些旨在最大化用户粘性的系统塑造。人工智能（AI）给了我们一个难得的机会，做得更好。
It is that when children are involved, safeguards must come before harm – not after. For years, social media platforms expanded with few meaningful protections for child safety, transparency, age-appropriate design, parental visibility or independent oversight. Only after mounting evidence of harm did policymakers begin responding in earnest. By then, millions of children had already been shaped by systems designed to maximize engagement.
斯塔默称：英国将禁止16岁以下儿童使用TikTok、YouTube等社交媒体应用 人工智能在教育领域蕴藏着非凡潜力。它能提供个性化支持、拓展教师能力、扩大受教育机会，并帮助学生以全新方式学习。但它也可能塑造儿童的思维、写作、推理及认知世界的方式。正因如此，这一刻至关重要。与社交媒体不同，AI带着后见之明而来。我们已深知，当强大技术在社会尚未为安全、问责与透明度确立明确规则前，便深度嵌入儿童生活，会发生什么。我们不应重蹈覆辙。
Artificial intelligence (AI) gives us a rare chance to do better. UK TO BAN TIKTOK, YOUTUBE, OTHER SOCIAL MEDIA APPS FOR CHILDREN UNDER 16, STARMER SAYS AI has extraordinary potential in education. It can personalize support, extend teacher capacity, widen access and help students learn in new ways. But it can also shape how children think, write, reason and engage with the world. That is why this moment matters. Unlike social media, AI arrives with the benefit of hindsight. We already know what happens when powerful technologies become deeply embedded in the lives of children before society establishes clear rules for safety, accountability and transparency.
记者手记：立法者纠结AI能否在美国课堂“及格” 一个负责任的教育AI框架，应以一个简单原则为起点：人工智能必须增强人类学习，而非取代人类或其判断。
We should not make the same mistake twice.
教师必须继续掌控教学、评分与课程。家长必须继续作为子女教育的主要决策者，并对AI如何与孩子互动、收集何种信息、如何生成推荐拥有实质性知情权。而学生必须持续培养构成真实学习的习惯：推理、写作、辨析、坚持与批判性思维。
REPORTER'S NOTEBOOK: LAWMAKERS WRESTLE OVER WHETHER AI CAN MAKE THE GRADE IN AMERICA'S CLASSROOMS A responsible framework for AI in education should begin with a simple principle: artificial intelligence must enhance human learning, not replace humans or their judgment. Teachers must remain in charge of instruction, grading and curriculum. Parents must remain the primary decision-makers in their children’s education, with meaningful visibility into how AI interacts with their child, what information it collects, and how recommendations are generated.
威廉·贝内特、约翰·哈格：人工智能可以模仿教师，却无法引导灵魂。这正是为什么这场讨论不能仅仅局限于人工智能本身，还必须涉及伦理智能。
And students must continue developing the habits that define real learning: reasoning, writing, discernment, persistence and critical thought. WILLIAM BENNETT, JON HAGE: AI CAN MIMIC A TEACHER, BUT IT CANNOT SHEPHERD A SOUL That is also why this conversation must be about more than artificial intelligence alone.
人工智能可以让学习更快、更个性化、更高效。而伦理智能则能确保其保持人性。它会提出技术自身无法回答的问题：什么是真理？什么是善？什么是公平？责任要求我们做什么？这个工具正在帮助孩子成为怎样的人？
It must also be about Ethical Intelligence. AI can make learning faster, more personalized and more efficient. Ethical Intelligence ensures it remains human. It asks the questions technology cannot answer on its own: What is true? What is good? What is fair? What does responsibility require?
当人工智能测试造成损害时，我们需要更强大的保障机制和真正的问责制度。这种区别至关重要。缺乏伦理的智能可能会为了便利而优化，同时削弱人的塑造过程。它可能提供答案却缺乏判断，带来速度却缺少智慧，实现个性化却忽视品格培养。在教育领域，这并非进步。
What kind of person is this tool helping a child become? WHEN AI TESTS CAUSE DAMAGE, WE NEED STRONGER SAFEGUARDS AND REAL ACCOUNTABILITY That distinction is essential. Intelligence without ethics can optimize for convenience while weakening formation. It can deliver answers without judgment, speed without wisdom, and personalization without character. In education, that is not progress.
这意味着，面向学生的AI在广泛部署之前，必须满足明确的标准。这些标准应包括：明确说明AI的使用方式和范围；强有力的学生隐私保护；限制数据收集与商业化；保留教师在评分和教学方面的权威；禁止操控、行为画像和隐性影响；提供适合年龄段的输出内容；以及确保家长能够知情、参与并监督。
That means student-facing AI should meet clear standards before it is widely deployed. Those standards should include transparency about where and how AI is used; strong protections for student privacy; limits on data collection and commercialization; preservation of teacher authority over grading and instruction; prohibitions on manipulation, behavioral profiling and hidden influence; age-appropriate outputs; and meaningful parental notice, access and oversight.
奥巴马批评特朗普的人工智能处理方式，并升级警告，呼吁加强监管。同样重要的是，教育领域的AI在面向数百万儿童之前，必须同时证明其安全性和有效性。课堂试点测试应当成为常态，而非例外。“先部署，再监管”是社交媒体时代的典型错误。教育领域不能重蹈覆辙。然而，最深层的保障措施并非技术性的，而是哲学性的。
OBAMA CRITICIZES TRUMP ON APPROACH TO AI, ESCALATES WARNING FOR STRONGER REGULATION Just as important, educational AI should demonstrate both safety and effectiveness before it reaches millions of children. Classroom-based pilot testing should be the norm, not the exception. "Deploy first and regulate later" was the defining mistake of the social media era. Education cannot afford to repeat it. The deepest safeguard, however, is philosophical, not technical. CHRISTIAN UNIVERSITY AIMS TO PRODUCE 'IRREPLACEABLE' GRADUATES AS AI RESHAPES HIGHER EDUCATION Artificial intelligence should function as a tutor, coach, research assistant and productivity tool — not as a substitute for the intellectual and moral development of children.
基督教大学旨在培养“不可替代”的毕业生，以应对人工智能重塑高等教育的挑战人工智能应作为导师、教练、研究助手和生产力工具发挥作用——而非取代儿童的智力与道德发展。教育不仅仅是获取答案，而是学习如何思考、如何判断、如何应对复杂性，以及如何在知识与品格上共同成长。韧性、智慧、创造力与信念，都是通过努力、反思与奋斗形成的。人工智能可以支持这一过程，但绝不能短路这一过程。
Education is not simply about getting answers. It is about learning how to think, how to judge, how to wrestle with complexity and how to grow in both knowledge and character. Resilience, wisdom, creativity and conviction are formed through effort, reflection and struggle. AI can support that process.
每项变革性技术都带来希望与风险并存。互联网扩大了知识获取渠道，同时也引入新的脆弱性。社交媒体连接了数十亿人，却也造成严重危害，尤其是对年轻用户。人工智能可能成为教育史上最有益的创新之一——但前提是必须有一个保护儿童、维护教育真正目的的治理框架。
It should never short-circuit it. Every transformative technology brings both promise and risk. The internet expanded access to knowledge while introducing new vulnerabilities. Social media connected billions of people while contributing to serious harms, especially for young users.
Meta新推青少年限制措施对年轻人意味着什么？这是政策制定者难得面临的机遇。 数字时代以来，领导人首次有机会在变革性技术全面嵌入儿童日常生活前，建立有意义的保障措施。通过坚持透明度、家长权利、教师权威、学生隐私、适龄设计、独立测试与教育问责，各州可以在鼓励创新的同时明确：技术必须服务于教育——而非相反。
AI may become one of the most beneficial innovations in the history of education — but only if it is governed by a framework that protects children while preserving the true purpose of schooling. WHAT META'S NEW TEEN RESTRICTIONS MEAN FOR YOUNG PEOPLE This is the rare opportunity now before policymakers. For the first time in the digital age, leaders can establish meaningful safeguards before a transformative technology becomes fully embedded in the daily lives of children. By insisting on transparency, parental rights, teacher authority, student privacy, age-appropriate design, independent testing and educational accountability, states can encourage innovation while making clear that technology must serve education — not the other way around.
社交媒体展示了创新超越责任的后果。人工智能提供了一个难得的机会来扭转这一教训。在人工智能重塑童年之前，我们必须确保“伦理智能”塑造人工智能。教育的未来、下一代的品格，正系于此。
Social media showed what happens when innovation outpaces responsibility. AI presents a rare opportunity to reverse that lesson. Before artificial intelligence reshapes childhood, we must ensure that Ethical Intelligence shapes artificial intelligence. The future of education, and the character of the next generation, depends on it.
埃德·波祖利是教育法领域的专家，同时也是特里普·斯科特律师事务所的首席执行官。
Ed Pozzuoli is an expert in education law and CEO of Tripp Scott Law Firm.