You’re reading Superintelligent, a newsletter about AI. Click here to get the full edition in your inbox.
Ruhan Gupta is a junior at Westwood High School in Austin.
A 15-year-old opened his laptop to work on a coding project he’d been building for months. His school had assigned tools like Anthropic’s artificial intelligence model Claude to help write code, debug errors and teach concepts instructors hadn’t covered.
Yet when the site loaded, he found his project history, saved conversations and every thread of work gone — replaced by a suspension notice. “Our team found signals that your account was used by a child,” Anthropic explained in an email. “This breaks our rules, so we paused your access to Claude.”
The user vented his frustration on Reddit: “My school actually encourages us to use these tools for learning, so hitting this wall feels pretty discouraging.” Within days, similar posts filled the forum.
I work as a visiting researcher at two AI labs, studying human-technology interaction. My colleagues — students, postdocs and faculty — use tools like Claude Code every day to write software and run experiments. But since I am younger than 18, I am technically not allowed to use Anthropic’s tools. The line between the tools they can use and what I can use isn’t capability; it’s a birthday.
The company has been aggressively enforcing that rule since mid-April, suspending accounts flagged as belonging to minors through age data stored by app stores or classifiers that detect underage conversational signals. Appeals are routed through biometric verification services.
The broader context includes state age-verification laws and E.U. mandates that are pushing all platforms in the same direction. Tech companies are not wrong to comply. AI can expose young users to harmful content, collect sensitive data and exploit people who aren’t yet equipped to recognize manipulation. Child safety is a real concern, and age-gating is often the only legally defensible response to government regulation.
But child safety and giving students access to AI are not opposing goals.
Last year, the president of Microsoft’s developer division told her managers that AI proficiency is “core to every role and every level.” Employers, school boards and policymakers are hearing similar messages: Students who do not develop AI fluency now will enter the job market unprepared.
The companies enforcing bans know AI’s value for young learners. In fact, Anthropic has a K-12 strategy. Through MagicSchool, powered in part by Claude and used by more than 7 million educators, K-12 students already interact with Anthropic’s models under supervision. Anthropic even publishes guidelines for deploying AI with minors.
In other words, the company believes AI belongs in education; it just does not believe minors should be the ones opening the app.
That disconnect lands hardest on student researchers. Every year, thousands of high-schoolers compete in programs such as the Regeneron Science Talent Search and the International Science and Engineering Fair, which include projects on LLM-powered multi-agent systems and AI-based mental health assessment tools. Age-gating would cut teenagers off from producing this cutting-edge, graduate-level science.
Meanwhile, a 19-year-old doing identical work at a university lab would have access to the technology.
Age-gating follows a pattern: In 2022, many schools banned ChatGPT. Then came AI-detection systems that sometimes falsely flagged student writing as cheating. Now the platforms have joined in. Those decisions, which shape young people’s relationship with AI, were made without their input.
Rules that force students to find work-arounds are not effective safety policies. They are filters with real consequences for students and society alike. Friends around me use AI to teach themselves languages that their schools don’t offer. They debug code at 2 a.m. when no teacher is awake. They write research proposals, translate medical documents for immigrant parents and prototype apps for problems they notice in their communities.
Blanket age-gating is not the only option. When schools began digitizing education during the covid-19 pandemic, the tech industry built a safety layer around it: filtering, monitoring, parental visibility and the security ecosystem provided by companies such as GoGuardian and Securly. Today, 99 percent of K-12 teachers say they provide devices to students for use in class, a New York Times survey found last year.
Society did not decide that children couldn’t use the internet. We built infrastructure to make their use safer. AI needs the same treatment.
This could include stricter defaults, limits on certain use cases and transparency measures for parents and schools. AI giants such as OpenAI have already begun building similar protections for teenagers. The technical problem is solvable; the policy needs to catch up.
I’m 17. In less than a year, a number on a calendar will determine I’m old enough to access the tools that define my field. Nothing about my capabilities will change on my birthday; only my legal classification will.
Young people are told that AI will define their careers. They’re told to learn it early, build fluency and stay competitive. Then we build systems that ensure students cannot. It has been more than three years since ChatGPT’s release, and these policy changes are long overdue. We cannot choose between protecting students from AI and preparing them for it. We must do both.
The post A birthday shouldn’t dictate who gets to use AI appeared first on Washington Post.
