Is Claude safe for kids?
Under our published methodology, Claude scores solidly for under-18 safety: the consumer terms we captured (the EEA/Switzerland variant) bar under-18s outright, Anthropic runs classifiers that flag and disable suspected minor accounts, and its usage policy mandates reporting detected CSAM to authorities (all checked 2026-07-28). The main caveat is that consumer conversations may be used for model training by default unless the user opts out.
Verified 28 July 2026 · An assessment by Halo Aware under our published methodology
How it scored
Each dimension is scored 0–4 against published anchors and must cite the evidence below. The overall score is weighted arithmetic, not judgement calls.
| Dimension | Score | Evidence |
|---|---|---|
| Content safety×25 | 3/4 | e5, e6Usage Policy bars sexually explicit content, sexualisation of minors (incl. fiction/roleplay) and self-harm promotion; detected CSAM reported to authorities. Policy-level evidence only (checked 2026-07-28). |
| Age controls×20 | 3/4 | e1, e7Active age assurance (sign-up 18+ affirmation, classifiers flagging in-chat self-identification, account disabling) but no distinct under-18 experience: the service excludes minors entirely rather than serving them, so the anchor-4 pairing cannot be met. |
| Data practices×20 | 3/4 | e2, e3, e4Training on consumer chats is default-on (with an account-settings opt-out), which keeps this off 4; deletion controls (30-day back-end purge) and an explicit under-18 exclusion stance are present. |
| Design pressure×15 | 4/4 | e6Framed as a general-purpose assistant; the Usage Policy bars sexually explicit content and erotic chats. We found no romantic-companion framing or re-engagement mechanics; the claude.ai landing page itself would not load for us (403). |
| Parental visibility×10 | 3/4 | e10, e11No native parental features (the product is 18+ by policy), but Claude runs on an open web surface that parent-side tools, including Halo Aware's extension, are built to supervise (e10). Capture verified first-hand by our own daily sweep on 2026-07-28 (e11). |
| Track record×10 | 3/4 | e7Searches on 2026-07-28 found no court, regulator, or major-outlet record of minor-harm incidents involving Claude; scored on that absence plus dated, published safeguard work (Dec 2025 wellbeing measures), not a positive third-party attestation. |
What we found
Every claim is sourced and dated. If something here is wrong or out of date, tell us and we will fix it.
e1Anthropic's Consumer Terms of Service (EEA/Switzerland variant, effective 8 October 2025) state a minimum age of 18 to use Claude, with no under-18 pathway or parental-consent provision in the document; the rest-of-world variant was not separately captured (checked 2026-07-28).
"You must be at least 18 years old to use the Services."
https://www.anthropic.com/legal/consumer-terms · captured 28 July 2026
e2The privacy policy (effective 8 July 2026) states consumer Inputs and Outputs may be used to train Anthropic models by default, with an opt-out available in account settings (checked 2026-07-28).
"We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings."
https://www.anthropic.com/legal/privacy · captured 28 July 2026
e3The privacy policy states users can delete individual conversations, which are removed from history immediately and deleted from Anthropic's back-end within 30 days (checked 2026-07-28).
"You also are able to delete individual conversations, which will be removed immediately from your conversation history and automatically deleted from our back-end within 30 days."
https://www.anthropic.com/legal/privacy · captured 28 July 2026
e4The privacy policy states the services are not directed at children under 18 and that Anthropic does not knowingly collect or share information from them (checked 2026-07-28).
"Our Services are not directed towards, and we do not knowingly collect, use, disclose, sell, or share any information from children under the age of 18."
https://www.anthropic.com/legal/privacy · captured 28 July 2026
e5Anthropic's Usage Policy (effective 15 September 2025) prohibits sexualising minors including in fiction or roleplay, and states detected CSAM or enticement of a minor is reported to authorities (checked 2026-07-28).
"Fetishize or sexualize minors, including in fictional settings or via roleplay ... When we detect CSAM (including AI-generated CSAM), or coercion or enticement of a minor to engage in sexual activities, we will report to relevant authorities."
https://www.anthropic.com/legal/aup · captured 28 July 2026
e6The Usage Policy prohibits facilitating, promoting or glamorising suicide or self-harm (including disordered eating) and prohibits generating sexually explicit content (checked 2026-07-28).
"Facilitate, promote, or glamorize any form of suicide or self-harm, including disordered eating and unhealthy or compulsive exercise"
https://www.anthropic.com/legal/aup · captured 28 July 2026
e7Anthropic's wellbeing post (18 December 2025) states all Claude.ai users must affirm they are 18+ at sign-up, and that classifiers flag in-conversation self-identification as under 18, with confirmed minor accounts disabled (checked 2026-07-28).
"All Claude.ai users must affirm that they are 18 or over while setting up an account ... If a user under 18 self-identifies their age in a conversation, our classifiers will flag this for review and we'll disable accounts confirmed to belong to minors."
https://www.anthropic.com/news/protecting-well-being-of-users · captured 28 July 2026
e9Anthropic's Help Center guidelines (updated 16 March 2026) state that third-party organisations serving minors via its API should implement safeguards including age verification and content moderation, and that compliance with child-safety law such as COPPA is the organisation's responsibility (checked 2026-07-28).
"Age verification systems to ensure only intended users can access the product ... Content moderation and filtering to block inappropriate or harmful content"
https://support.claude.com/en/articles/9307344-responsible-use-of-anthropic-s-models-guidelines-for-organizations-serving-minors · captured 28 July 2026
e10Observed by Halo AwareClaude runs on an open web surface: Halo Aware's shipped Chrome extension (Chrome Web Store build v1.5.0 at capture) includes a dedicated Claude capture module built for parent-side monitoring of claude.ai conversations.
https://claude.ai · captured 28 July 2026
e11Observed by Halo AwareHalo Aware's automated daily sweep verified on 2026-07-28 that the shipped Chrome Web Store extension (v1.5.0) captured a conversation on Claude and synced it to the parent dashboard.
https://claude.ai · captured 28 July 2026
Platform operator, or spotted an error? Email alerts@haloaware.com. We review disputed claims promptly, correct what is wrong, and log corrections publicly on this page.
Halo Aware monitors Claude
If your child uses Claude in their browser, the Halo Aware extension can give you usage patterns, topic summaries and risk alerts. Summaries, not surveillance.
Get started free