Anthropic Restricts Claude Access to Users Aged 18 and Over

TL;DR

• Anthropic updated its Terms of Service to restrict Claude access strictly to users aged 18 and older [^1].
• The change introduces age assurance mechanisms, cutting off high school students and underage developers who relied on the platform [^1][^2].

Background

AI safety and child protection regulations have intensified globally, putting generative AI providers under strict legal scrutiny [^1][^2]. Laws like the UK's Online Safety Act and California's Age-Appropriate Design Code impose heavy liability on platforms that serve underage users without stringent parental controls or age verification [^2]. While frontier AI models excel at coding assistance and academic research, operating services for minors introduces liability surrounding data privacy, unexpected content output, and algorithmic safety standards [^1].

What happened

Anthropic officially revised its Terms of Service and user policy guidelines, restricting all Claude access to individuals who are at least 18 years old [^1]. Previously, users aged 13 to 17 could access Claude with parental or legal guardian consent [^1]. Under the updated framework, accounts identified as belonging to minors will face suspension, and new account creations now require compliance with automated age assurance protocols [^1][^2].

To enforce this restriction, Anthropic is deploying tiered age verification methods [^1]. The system uses background signals, payment details, and identity verification checks to flag account profiles likely registered by minors [^1][^2]. If the system detects suspicious age indicators during registration or routine login flows, the user must present valid government identification or undergo third-party age verification before regaining access to the workspace [^1].

The policy shift applies across all consumer-facing Claude tiers, including the free web interface, Claude Pro subscriptions, and mobile apps [^1][^2]. Anthropic clarified that enterprise software deployments and API developers managing their own end-user authentication pipelines remain subject to separate commercial agreements, provided those third-party services implement appropriate compliance guardrails [^1].

Why it matters

This policy update signals a significant operational shift in how frontier AI labs manage legal risk and regulatory exposure. For years, major AI companies allowed teenager access under broad parental consent clauses, mirroring early social media policies. However, as international regulatory authorities increase financial penalties for mismanaging minor data or failing to prevent harmful model outputs, AI laboratories are deciding that serving minors directly is no longer commercially viable. Restricting access to adults simplifies compliance frameworks, shields companies from youth-focused safety lawsuits, and eliminates the need to build specialized filtering tiers for high school accounts.

However, the ban creates an immediate disruption for student programmers, educators, and young researchers. Claude has become a staple tool in high school computer science classrooms and extracurricular coding bootcamps due to its strong reasoning capabilities and helpful coding output. Excluding students under 18 pushes young developers toward alternative open-weight models or less regulated AI platforms. It also establishes a clear divide between enterprise API developers—who can still build tools for minors using Claude's backend if they assume legal liability—and direct consumer users who are locked out entirely.

Furthermore, enforcing age restrictions on digital platforms inherently requires expanding data collection practices. To verify whether a user is over 18, AI platforms must either partner with identity verification vendors, process biometric age estimation data, or analyze user behavioral patterns. This creates an awkward paradox: in the pursuit of protecting minor privacy and compliance, platform operators must collect more sensitive personal identity data from adult users. This dynamic highlights the growing friction between user privacy and strict regulatory enforcement across the AI ecosystem.

Practical example

Consider a 17-year-old high school senior named Liam who uses Claude Pro every afternoon to debug his AP Computer Science projects and draft code for a robotics competition team.

On Tuesday morning, Liam opens his browser and attempts to log into his Claude account to review a Python script. Instead of his workspace dashboard, a prompt notifies him that his account requires age verification to continue.

Because Liam is under 18, he cannot pass the automated verification check or upload adult identification documents. His subscription access freezes, locking him out of his saved conversation history and custom prompts.

To finish his project deadline, Liam must migrate his workflow to an open-weight model running locally on his laptop or rely on alternative platforms that still allow minor access with parental consent.

Related gear

We recommend this book because it explores the regulatory, ethical, and safety challenges that drive AI companies to restrict model access.