Claude is only available to people over 18 years
Anthropic’s policy that its Claude AI assistant is only for users over 18, enforced via third‑party age verification (Yoti), is prompting backlash over privacy, biometric data collection, and the broader trend toward ID‑gated internet services. Supporters frame the move as a pragmatic response to legal risk and emerging child‑safety laws, arguing that minors create outsized liability for little revenue, while critics see “compliance theater” that offloads responsibility to parents and chips away at anonymity without meaningfully protecting children. Many expect the policy to push younger users and privacy‑conscious adults toward local or open‑weights models and non‑US providers, highlighting a growing divide between regulated commercial AI and self‑hosted alternatives.
Policy timing and scope
- Many note the support article is dated May 2026 and earlier ToS already restricted minors; this is not a brand‑new change.
- Users report accounts being locked after mentioning being under 18, suggesting automated “minor detection” was already live.
- The policy covers direct Claude use; it’s unclear how consistently it applies to third‑party wrappers and API‑based products.
Motivations and legal/regulatory context
- A common view: this is primarily risk and liability management, not “data grab.”
- Commenters cite COPPA, UK Age Appropriate Design Code, EU GDPR, and California’s SB 1119 (“Adam’s Law”) as regulatory pressure making minors disproportionately risky and unprofitable.
- Some see it as “compliance theater” to avoid building robust child‑safety guardrails.
Privacy, surveillance, and ID verification concerns
- Strong resistance to giving government IDs or biometrics to Yoti or any KYC vendor, especially after widely discussed ID‑verification data breaches.
- Skepticism that “we only get a pass/fail result” and “data deleted immediately” are enforceable; SOC2 is viewed as weak reassurance.
- Several see age‑gating as a step toward full online identity binding and “surveillance society,” potentially enabling state or corporate overreach.
Impact on minors, education, and parental roles
- Many argue it unfairly excludes teens from an important learning and creativity tool, analogous to banning kids from calculators or the early web.
- Others counter that minors have special vulnerabilities (impulse control, susceptibility to self‑harm content), and the upside isn’t worth the legal and ethical risk.
- Debate over whether companies or parents should decide children’s access; some parents oppose any centralized age checks, others welcome friction as a safety measure.
- Suggestions include OS‑level or device‑level “child lock” flags and privacy‑preserving age proofs (e.g., ZKPs, bank‑mediated attestations), but feasibility and abuse potential are disputed.
Workarounds and ecosystem effects
- Many say they’ll cancel Claude and use local or open‑weight models (often Chinese) or other hosted providers without ID checks.
- Some predict this will push younger users and schools toward open‑source models and alternative ecosystems, possibly changing the “industry standard.”
Broader debate about AI harms
- Opinions diverge on whether LLMs are “brain‑rotting fast food,” a powerful tutor, or both.
- Several argue social media is a far larger proven harm to minors, and see age‑gating LLMs as misdirected or hypocritical.