Risk Score
0 = very fair · 100 = very risky
Summary
This is Anthropic's Terms of Service for Claude.ai and Claude Pro, governing individual use of their AI assistant products. The document is moderately balanced for a SaaS AI product, containing some standard industry clauses alongside a few more aggressive provisions. Notably, users retain ownership of their inputs and receive assignment of outputs, and the document provides clearer-than-average cancellation procedures. However, there are significant liability caps, broad indemnification requirements, unilateral rights to modify or terminate services, and an opt-out (rather than opt-in) model training data use provision that warrant attention.
Flagged Clauses
By default, your conversations (inputs and outputs) can be used to train Anthropic's AI models. You have to actively find and toggle an opt-out setting to prevent this — it is not off by default.
“We may use Materials to provide, maintain, and improve the Services and to develop other products and services, including training our models, unless you opt out of training through your account settings.”
Even if you opt out of training, Anthropic can still use your conversations for model training if you give feedback (e.g., thumbs up/down) or if your conversation is flagged by safety systems. This creates exceptions that partially undermine the opt-out.
“Even if you opt out, we will use Materials for model training when: (1) you provide Feedback to us regarding any Materials, or (2) your Materials are flagged for safety review.”
You keep ownership of what you type in, and Anthropic assigns any rights they have in AI-generated responses back to you. This is more user-friendly than many AI service terms, though the phrase 'if any' signals uncertainty about whether AI outputs are legally copyrightable.
“As between you and Anthropic, and to the extent permitted by applicable law, you retain any right, title, and interest that you have in the Inputs you submit. Subject to your compliance with our Terms, we assign to you all of our right, title, and interest—if any—in Outputs.”
Anthropic can change, pause, or shut down the service entirely at any time with no prior warning to you, and they state they will not be held liable for doing so.
“We reserve the right to modify, suspend, or discontinue the Services or your access to the Services, in whole or in part, at any time without notice to you.”
Anthropic can change the rules at any time. Continuing to use the service after changes are posted counts as your agreement to those changes, even if you didn't actively read or approve them.
“We may revise and update these Terms at our discretion... If you continue to access the Services after we post the updated Terms on Anthropic's website or otherwise give you notice of Terms changes, then you agree to the updated Terms.”
Anthropic can terminate your account without warning if they believe (not necessarily prove) you broke the rules. Upon termination, Materials and account data may be deleted at Anthropic's option.
“We may suspend or terminate your access to the Services (including any Subscriptions) at any time without notice to you if we believe that you have breached these Terms... we may also terminate your Account if you have been inactive for over a year.”
If your account is terminated — for any reason — Anthropic may permanently delete all of your conversation history and associated data. There is no stated obligation to give you the chance to export your data first.
“Upon termination of these Terms, a Subscription, or your access to the Services, we may at our option delete any Materials or other data associated with your Account.”
Most payments are non-refundable. If you are terminated for a Terms violation, no refund is owed. The primary refund protection only applies in specific jurisdictions (Brazil, Mexico, South Korea, Taiwan) or when Anthropic terminates for non-violation reasons.
“Except as expressly provided in these Terms or where required by law, all payments are non-refundable.”
Subscriptions automatically renew and your payment method is charged unless you cancel at least 24 hours before the renewal date. There is no mention of an advance reminder notice before renewal charges occur.
“Your Subscription will last for the Initial Term and will automatically renew... until you cancel. To avoid renewal and charges for the next Renewal Term, cancel your subscription at least 24 hours before the last day of the Initial Term or any Renewal Term.”
No matter what goes wrong — including data loss, service failures, or harm caused by AI outputs — the maximum Anthropic will pay you is either what you paid in the last 6 months or $100, whichever is higher. For many users, this is a very low ceiling.
“THE ANTHROPIC PARTIES' TOTAL AGGREGATE LIABILITY TO YOU FOR ALL DAMAGES... WILL NOT EXCEED THE GREATER OF THE AMOUNT YOU PAID TO US FOR ACCESS OR USE OF THE SERVICES IN THE SIX MONTHS PRECEDING... AND $100.”
If Anthropic faces any legal claims related to your use of the service — even allegations, not proven violations — you agree to cover Anthropic's legal costs and any damages. This is a broad obligation that could be financially significant.
“YOU AGREE TO INDEMNIFY AND HOLD HARMLESS THE ANTHROPIC PARTIES FROM AND AGAINST ANY AND ALL LIABILITIES, CLAIMS, DAMAGES, EXPENSES (INCLUDING REASONABLE ATTORNEYS' FEES AND COSTS)... ARISING OUT OF OR RELATED TO YOUR BREACH OR ALLEGED BREACH OF THESE TERMS.”
Features you pay for in a subscription (like higher usage limits or certain capabilities) can be changed or removed at any time. The service you subscribed to may not remain the same throughout your subscription period.
“We may change the content, features, and other services from time to time, and we do not guarantee that any particular piece of content, feature, or other service will always be available through the Services.”
Anthropic can voluntarily share your conversation content with law enforcement at their own discretion, not just when legally compelled to do so by a court order or subpoena.
“We reserve the right, at our sole discretion, to report information from or about you, including but not limited to Inputs, Outputs, or Actions to law enforcement.”
If you sign up with a work email, your employer could potentially see your conversations. Anthropic may rely on your employer to notify you of this rather than notifying you directly.
“If you use an email address owned by your employer or another organization, your Account may be linked to the organization's Anthropic enterprise account, and the organization's administrator may be able to monitor and control the Account, including having access to Materials.”
Cancellation is possible through the customer portal or by email, which is relatively straightforward. Multiple cancellation pathways are provided.
“If you subscribed via our website, you may cancel your Subscription for any reason by using a method we may provide to you through our products—for example, for Claude Pro, in your customer portal—or by notifying us at support@anthropic.com.”
Missing Protections
- No explicit arbitration clause visible in the provided text — but the document appears to be cut off before Section 13 is complete, so mandatory arbitration or class action waiver provisions may exist in the full document
- No stated obligation to provide advance notice before automatic subscription renewal charges
- No data portability or export rights — no provision allowing users to download their conversation history before account termination
- No explicit data retention schedule specifying how long user data (including conversations) is kept after account closure
- No explicit opt-in consent mechanism for the model training data use — it is opt-out by default rather than opt-in
- No stated process or timeline for responding to account termination appeals
- No explicit GDPR or CCPA rights section in the provided Terms text (though this may be in the referenced Privacy Policy)
- No service level agreement (SLA) or uptime guarantee of any kind
- No explicit prohibition on selling user data to third parties (this may be addressed in the Privacy Policy, but is not stated here)
Fair Terms
- Users retain ownership of their inputs AND receive an assignment of Anthropic's rights in AI-generated outputs — more protective than many competing AI services that claim broad licenses
- Clear opt-out mechanism for model training via account settings is explicitly mentioned
- Fee increase requires at least 30 days advance notice before taking effect, and users can cancel to avoid the increase
- Fee increases cannot take effect during a current billing term — only at renewal
- Pro-rata refund is provided if Anthropic terminates a subscription for reasons other than user violations
- 7-day cancellation window with full refund rights provided for residents of Brazil, Mexico, South Korea, and Taiwan
- Inactive account termination includes advance notice before deletion
- Explicit acknowledgment that AI outputs may be inaccurate and users should independently verify them — sets appropriate expectations
- Minimum age set at 18 (or higher local age of consent), providing a baseline protection against minor account creation
Document information only — not legal advice.