Is Claude safe? Age rules, content limits, privacy and legal actions

The short answer

It depends on who is using it and what they share. As of October 22, 2026, Anthropic's consumer terms say users must be at least 18. Anthropic's help center says it may disable accounts its systems flag as possibly belonging to someone under 18 and ask for an age check. We did not find an Anthropic page describing a teen mode or parental controls.

On privacy, Anthropic's privacy policy says it may use chats to train its models unless you opt out, and its privacy center says chats kept for training may be stored in de-identified form for up to five years. Anthropic says designated trust and safety staff can review conversations to enforce its rules.

Anthropic was among 13 companies that state attorneys general wrote to in August and December 2025. The Associated Press reported on September 30, 2026 that the FTC had confirmed an investigation of Anthropic, OpenAI and other AI companies over possible risks to consumers. A proposed class action that Courthouse News reported was filed in September 2026 alleges that Anthropic's ID-and-selfie checks violated an Illinois biometric privacy law. These are letters, an investigation and allegations, not findings.

Who can use Claude and how age is checked

• Minimum age: Anthropic's Consumer Terms of Service, effective October 8, 2025, say you must be at least 18, or older if your location sets a higher minimum age to consent to the service.
• Children's data: Anthropic's privacy policy, effective September 10, 2026, says its services are not directed to children, that it does not knowingly collect information from anyone under 18, and that it has measures to detect and remove children from its services.
• Sign-up and flagging: in a December 18, 2025 post, Anthropic said users affirm they are 18 or older when they sign up, that conversations in which a user says they are under 18 are flagged for review, and that accounts confirmed to belong to minors are disabled. It said it was building a classifier to catch subtler signs of underage use.
• Age assurance: Anthropic's help center article on age assurance, dated May 18, 2026, says an account may be disabled if its safety systems detect signs the user may be under 18, and the user is then asked to verify their age through Yoti, a third-party provider, using a selfie-based age estimate, a photo ID or the Yoti Digital ID app. Anthropic says it receives only a pass or fail result and that Yoti deletes the images after the check.
• App store age signals: a separate help article dated March 16, 2026 says some U.S. states now require app stores to verify ages and share that information with developers, and that a user whom the App Store or Google Play has identified as under 18 may be unable to create an account or sign in through the Claude mobile app. Our guide to app store age laws explains those state laws.
• App store rating: as of October 22, 2026, Apple's App Store lists the Claude app as 18+, with infrequent mature or suggestive themes and infrequent sexual content or nudity among its content descriptors, and lists age assurance as an in-app control.
• Parental controls: we did not find an Anthropic page describing parental controls, a teen mode or teen-specific protections.

Content rules and crisis features

Anthropic published an updated Usage Policy on October 8, 2026 that takes effect November 12, 2026. The policy on its site prohibits, among other things:

• creating or distributing child sexual abuse material, including AI-generated material, and sexualizing minors, including in fiction or role-play; Anthropic says it reports detected child sexual abuse material to the relevant authorities;
• promoting, glamorizing or facilitating suicide or self-harm, and content that promotes eating disorders;
• sexually explicit content, including erotic chats;
• creating or threatening to share nonconsensual intimate images, including AI-generated ones, and impersonating real people.

Anthropic's Usage Policy says it may warn users or limit, suspend or end their access if it suspects a violation.

Crisis features: in its December 18, 2025 post, Anthropic said a classifier on Claude.ai looks for signs of suicidal ideation, and that a banner then points users to a helpline or other country-specific resources supplied by ThroughLine. Anthropic said Claude is not a substitute for professional advice or medical care.

Accuracy: Anthropic's consumer terms say outputs may be inaccurate and that users should not rely on them without independently checking them.

Privacy: training, staff access, retention and deletion

These rules cover Claude's consumer plans (Free, Pro and Max). Anthropic says work, education, government and API accounts are handled under different terms.

• Training: Anthropic's privacy policy says it may use your inputs and outputs to train its models unless you opt out in your account settings. In an August 28, 2025 announcement, Anthropic said consumer users would be asked to choose whether their chats can be used, with a decision required by October 8, 2025, and that the change applies to new or resumed chats.
• Opting out: Anthropic's help center says to go to Settings, then Privacy, and switch off the model-improvement toggle. It says turning it off stops future training on your chats but does not remove data already used in training that has started or in models already trained.
• Exceptions: Anthropic's privacy policy says conversations flagged for safety review, and conversations you send as feedback (such as a thumbs up or down), may still be used even if you opt out.
• Retention: Anthropic's privacy center, updated July 1, 2026, says chats kept for training may be stored in de-identified form for up to five years. Content flagged for Usage Policy violations may be kept for up to two years, and related safety classification scores for up to seven years. Feedback data is kept for five years.
• Staff access: Anthropic's privacy center says employees cannot access your conversations by default, except when you share them as feedback or when review is needed to enforce its Usage Policy, in which case designated trust and safety staff may see them.
• Incognito chats: Anthropic's help center says incognito chats are not saved to your chat history or memory and are not used for training, and are retained for 30 days by default.
• Deleting chats: Anthropic's privacy center says a deleted conversation disappears from your history right away and is deleted from its back-end systems within 30 days. It says data may be kept longer when needed to enforce its Usage Policy or comply with the law.
• Deleting your account: Anthropic's help center says to cancel any paid plan and wait for the billing period to end, then go to Settings, then Account, and select Delete Account. It says deletion is permanent.
• Identity checks: a help article dated August 11, 2026 says Claude may ask some users for a government photo ID and a live selfie, collected by Persona Identities, when they access certain capabilities or during platform integrity checks. Anthropic says it uses this data only to confirm identity and meet legal and safety obligations, and does not use it to train its models.
• Sale and advertising: Anthropic's privacy policy says it does not sell personal data as applicable laws define selling, and that users can opt out of sharing their personal data for targeted advertising that promotes Anthropic's products. As of October 22, 2026, the Claude app's App Store privacy label, which Anthropic supplies, lists contact info, identifiers, coarse location and product interaction as used for the developer's advertising or marketing, and lists no data used to track users.

Government letters and actions

These are letters and a reported investigation, not findings or enforcement cases.

• 44 attorneys general: on August 25, 2025, a bipartisan group co-led by Illinois, North Carolina, South Carolina and Tennessee wrote to 13 companies, including Anthropic, asking them to adopt safeguards that keep their AI products from sexualizing children.
• 42 attorneys general: in a letter announced December 10, 2025 and led by Pennsylvania, the group wrote to 13 AI companies, including Anthropic, asking for safety testing, recall procedures, clear consumer warnings and other safeguards, and asked the companies to commit to changes by January 16, 2026.
• FTC (reported): the Associated Press reported on September 30, 2026 that an FTC spokesperson confirmed the agency had opened an investigation into OpenAI, Anthropic and other AI companies over possible risks to consumers, and declined further comment. We did not find an FTC release about it as of October 22, 2026. An investigation is not a finding.

On the official sites we checked, we did not find a state or federal enforcement action against Anthropic over Claude's consumer safety or privacy practices as of October 22, 2026.

Lawsuits

Jose Enrique Ortiz Colon v. Anthropic, PBC is a proposed class action that Courthouse News reported on September 30, 2026 had been filed in San Francisco County Superior Court. The copy it published is dated September 29, 2026 and shows no case number. It brings two claims under the Illinois Biometric Information Privacy Act. The complaint claims Anthropic did not give written notice of how long it would keep face-geometry data that it alleges was created during Claude's ID-and-selfie verification, and did not publish a retention and destruction schedule for that data. An Anthropic spokesperson told Courthouse News the company was reviewing the complaint. These are allegations, not findings.

Authors and music publishers have also sued Anthropic in federal court in California alleging copyright infringement. In one authors' case, Bartz v. Anthropic PBC, the court granted final approval on July 20, 2026 of a $1.5 billion class settlement covering books in versions of the LibGen and PiLiMi datasets that, according to the order, Anthropic downloaded. A settlement is not a finding of liability.

Our chatbot lawsuits page tracks other suits involving AI chatbots.

Which laws may apply

None of the laws below names Claude. Whether Claude counts as a companion chatbot, an AI companion or a conversational AI service under a particular state law is a legal question. These laws set duties for chatbot operators:

• California SB 243, in effect since January 1, 2026: companion chatbot operators must keep a protocol for suicidal ideation, suicide and self-harm that refers users to crisis services, and must disclose that companion chatbots may not be suitable for some minors. The bill author's office says it gives families a private right of action.
• New York's AI companion law, in effect since November 5, 2025: operators must have a protocol to detect and respond to signs of suicidal ideation or self-harm and refer users to crisis services, and must tell users they are talking to AI at the start of a session and every three hours of continued use. The Attorney General enforces it.
• Idaho's Conversational AI Safety Act (S 1297), signed March 31, 2026 and effective July 1, 2027: operators of conversational AI services open to the public must disclose that the service is AI when people could think they are talking to a human, adopt a suicide-response protocol with crisis referrals, and meet extra duties for minor account holders. The law excludes some tools, and the Attorney General enforces it.

We are not saying Anthropic is covered by, or out of compliance with, any of these laws. It may have practices it has not published, and whether a law applies to it is a legal question this guide does not decide.