Anthropic
Anthropic collects extensive data about you by default — including your conversations, device location, advertising identifiers, and potentially biometric data — and uses your chats to train its AI unless you opt out, though that opt-out is riddled with exceptions, including that simply rating a response permanently re-enrolls that conversation in training. The company can voluntarily share your data with law enforcement on its own judgment, transfer your account to any third party without your consent, and change its terms at any time with your continued use counting as acceptance. If something goes wrong, Anthropic's liability to you is capped at just €100 or six months of fees, and a forced arbitration clause means you cannot sue in court or join a class action to challenge this. Business API customers receive meaningfully stronger protections than ordinary consumers, who bear the greatest exposure under these terms.
Last analyzed
Heads up: this analysis is AI-generated, can be incomplete or wrong, and is not legal advice. Use it as a starting point and read the original document before you agree — see our Terms.
Clause-by-clause breakdown
Every clause we flagged in Anthropic's agreement, with what it means for you.
Your chats train AI by default; opt-out has significant exceptions.
“We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings. Even if you opt-out, we will use Inputs and Outputs for model improvement when: (i) your conversations are flagged for safety review to improve our ability to detect harmful content, enforce our policies, or advance AI safety research, or (ii) you've explicitly reported the materials to us (for example via our feedback mechanisms).”
Quoted verbatim from the agreement What this means: Anthropic uses your conversations with Claude to train its AI models by default. You can opt out in account settings, but two large exceptions remain: if any conversation is flagged for safety review, or if you ever use the thumbs-up/thumbs-down feedback buttons, that content is used for training regardless of your opt-out. This means opting out does not fully protect your conversations from being used.
What you can do: Go to your account settings and opt out of model training. Avoid using the thumbs-up/thumbs-down feedback buttons if you want to minimize training use of your conversations.
Anthropic may collect your facial geometry and government ID images.
“Verification Data: In certain circumstances, we may ask you to verify your age or identity. If you choose to do so, data we will collect includes, depending on the method: an image of your government-issued identity document and the information appearing on it (such as your ID number and date of birth); your image in photo or video form, facial geometry templates (which may be considered 'biometric data' in some jurisdictions); and the result of the verification (for example, whether your age meets the applicable threshold).”
Quoted verbatim from the agreement What this means: Anthropic may collect biometric data including facial geometry templates and government ID images for identity or age verification. Biometric data is among the most sensitive categories of personal information, and the policy does not specify how long this data is retained or whether it is shared with third-party verification providers.
What you can do: Only complete identity verification if required; check the privacy policy of any third-party verification service Anthropic uses before submitting biometric data.
No specific retention periods are stated; data kept 'as reasonably necessary.'
“Anthropic retains your personal data for as long as reasonably necessary for the purposes and criteria outlined in this Privacy Policy and explained further in our privacy center. When the personal data collected is no longer required by us, we and our service providers will perform the necessary procedures for destroying, deleting, erasing, or converting it into an anonymous form as permitted or required under applicable laws.”
Quoted verbatim from the agreement What this means: The policy does not specify any concrete retention periods for most categories of personal data, only saying data is kept 'as long as reasonably necessary.' This vague standard gives Anthropic broad discretion to retain your data for extended periods. The only specific timeframe mentioned is 30 days for deleted conversations to be removed from back-end systems.
What you can do: Submit a data deletion request to [email protected] if you want your data removed; note that some data may be retained for legal or safety reasons.
Continuing to use Claude means you accept any updated terms automatically.
“By agreeing to these Terms, or in the absence of such agreement, by using the Services, you agree to be bound by them, including any changes made to them in accordance with the Terms.”
Quoted verbatim from the agreement What this means: Anthropic can change its Terms of Service and your continued use of the service counts as acceptance of those changes. You are not required to give new, affirmative consent to updated terms — simply continuing to use Claude binds you to whatever the new terms say.
What you can do: Monitor your email and the Anthropic website for policy change notifications; stop using the service before any change takes effect if you disagree with the new terms.
Connecting a health app shares your diagnoses and medications with Anthropic.
“We collect the following categories of consumer health data where you provide it to us directly, including if you elect to connect a third-party application to Claude.ai that contains your consumer health data: Information about your health-related conditions, symptoms, status, diagnoses, medical interventions, surgeries, procedures, testing, or treatments; Vital signs, bodily functions, symptoms, or health measurements that reveal your health status; Information about your use or purchase of prescribed medication; Information, including location or precise location information, that could identify your attempt to seek health care services or supplies; and Other information that may be used to infer or derive consumer health data. We use your consumer health data for the following purposes: To integrate your third party health application with Claude.ai so that you share your data with Claude to discuss; To prevent and investigate fraud, abuse, and violations of our Usage Policy, unlawful or criminal activity, unauthorized access to or use of personal data or Anthropic systems and networks, to protect our rights and the rights of others, and to meet legal, governmental and institutional policy obligations; In an aggregated form, to help us develop, test and improve our services.”
Quoted verbatim from the agreement What this means: If you connect a health app to Claude, Anthropic collects detailed health data including diagnoses, medications, vital signs, and location data that could reveal you sought healthcare. This data is also used in aggregated form to develop and improve Anthropic's services, meaning your sensitive health information contributes to product development.
What you can do: Avoid connecting third-party health apps to Claude unless necessary; revoke any existing health app integrations in your Settings.
Rating a response overrides your training opt-out permanently.
“Our use of Materials. 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, 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 to improve our ability to detect harmful content, enforce our policies, or advance our safety research.”
Quoted verbatim from the agreement What this means: Even after opting out of training, if you give any feedback (thumbs up/down) on a response, that entire conversation is stored and used for training. This means a simple click to rate a response permanently removes your opt-out protection for that conversation.
What you can do: Avoid using the feedback rating buttons if you have opted out of training and want to keep your conversations private from model training.
Anthropic can change the rules anytime, including pricing.
“Changes to these Terms. We may revise and update these Terms. For example, we may update these Terms (1) to reflect changes in our Services, like when we add or remove features or services, or update our pricing, (2) for security or legal reasons, or (3) to promote safety or prevent abuse. We will notify you of changes to the Terms which are reasonably likely to affect your use of the Services or legal rights. These changes will come into effect no less than 30 days from when we notify you, unless the change is due to a change in law or for security reasons (in which case we may need to change the Terms on shorter notice).”
Quoted verbatim from the agreement What this means: Anthropic can change these Terms at any time for broad reasons including pricing updates. While they promise 30 days' notice for material changes, changes due to 'security reasons' can take effect immediately with no advance notice. If you keep using the service after changes take effect, you are bound by the new terms.
What you can do: Watch for notification emails from Anthropic about Terms changes. If you disagree with a change, you can terminate your subscription before it takes effect and receive a pro-rata refund.
Anthropic's maximum payout to you is €100 or six months' fees.
“Limitation of liability. Except as otherwise set out in No Limitation above, our total liability to you for any loss or damage arising out of or in connection with these Terms, whether in contract (including under any indemnity), tort (including negligence) or otherwise will be limited to the greater of: (a) the amount you paid to us for access to or use of the Services in the six months prior to the event giving rise to the liability, and (b) €100.”
Quoted verbatim from the agreement What this means: If Anthropic causes you harm — for example, through a data breach, service failure, or negligent AI output — the most they will ever pay you is either what you paid them in the last six months or €100, whichever is higher. For a monthly subscriber paying €20/month, that cap could be as low as €120 for six months of payments.
What you can do: Be aware that you cannot recover significant damages from Anthropic even if their service causes you real harm. Do not rely on Claude for high-stakes decisions where errors could cause you financial or other serious loss.
Anthropic can voluntarily report your chats to law enforcement.
“Legal Compliance. We may comply with governmental, court, and law enforcement requests or requirements relating to provision or use of the Services, or to information provided to or collected under our Terms. We reserve the right, at our sole discretion, to report information from or about you, including but not limited to Inputs or Outputs, to law enforcement.”
Quoted verbatim from the agreement What this means: Anthropic can proactively share your conversations and personal information with law enforcement at their own discretion — not just when legally required. This means your chat history could be reported to police or government agencies without a court order and without notifying you.
What you can do: Do not share sensitive personal information in your Claude conversations that you would not want law enforcement to potentially see. Treat Claude conversations as potentially non-private.
Rating a response re-enrolls that conversation in training, ignoring your opt-out.
“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, 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 to improve our ability to detect harmful content, enforce our policies, or advance our safety research.”
Quoted verbatim from the agreement What this means: Even after opting out of training, giving a thumbs up or thumbs down on any response automatically re-enrolls that entire conversation into training data. This is a significant gotcha: a casual rating gesture overrides your deliberate opt-out decision for that conversation.
What you can do: Avoid using the thumbs up/thumbs down feedback buttons if you have opted out of training and want to keep your conversations out of training data.
Anthropic can voluntarily report your conversations to law enforcement anytime.
“Legal Compliance. We may comply with governmental, court, and law enforcement requests or requirements relating to provision or use of the Services, or to information provided to or collected under our Terms. We reserve the right, at our sole discretion, to report information from or about you, including but not limited to Inputs or Outputs, to law enforcement.”
Quoted verbatim from the agreement What this means: Anthropic can proactively share your conversations with law enforcement on its own initiative — not just in response to a court order or legal demand. The phrase 'at our sole discretion' means no legal process is required before Anthropic decides to report your inputs or outputs to authorities. This goes beyond standard legal compliance language.
What you can do: Treat all conversations with Claude as potentially disclosable to law enforcement; avoid sharing sensitive personal information you would not want authorities to see.
Anthropic may collect your face scan and government ID for age verification.
“Verification Data: In certain circumstances, we may ask you to verify your age or identity. If you choose to do so, data we will collect includes, depending on the method: an image of your government-issued identity document and the information appearing on it (such as your ID number and date of birth); your image in photo or video form, facial geometry templates (which may be considered 'biometric data' in some jurisdictions); and the result of the verification (for example, whether your age meets the applicable threshold).”
Quoted verbatim from the agreement What this means: If Anthropic asks you to verify your identity, they may collect a photo of your government ID, a photo or video of your face, and facial geometry templates (biometric data). The policy does not specify how long this sensitive biometric data is retained or whether it is deleted after verification is complete.
What you can do: Only complete identity verification if required to access the service; check your privacy settings afterward to see if any deletion options are available for verification data.
You waive your right to sue in court or join a class action.
“EACH PARTY AGREES THEY ARE WAIVING THE RIGHT TO A TRIAL BY JURY, AND THE RIGHT TO JOIN AND PARTICIPATE IN A CLASS ACTION, TO THE FULLEST EXTENT PERMITTED UNDER THE LAW IN CONNECTION WITH THESE TERMS.”
Quoted verbatim from the agreement What this means: Both the Commercial Terms and the K-12 Terms require all disputes to go through binding arbitration and explicitly strip away your right to sue in court or join a class-action lawsuit. This means if Anthropic harms many users in the same way, each person must fight their case alone in a private arbitration process rather than banding together. There is no opt-out window mentioned in either document.
What you can do: If you are a commercial or K-12 customer, be aware you are giving up your right to a jury trial and class action before signing. Consult a lawyer if the stakes are high enough to matter.
Opting out of AI training doesn't fully protect your conversations.
“We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings. Even if you opt-out, we will use Inputs and Outputs for model improvement when: (i) your conversations are flagged for safety review to improve our ability to detect harmful content, enforce our policies, or advance AI safety research, or (ii) you've explicitly reported the materials to us (for example via our feedback mechanisms).”
Quoted verbatim from the agreement What this means: Even after you opt out of model training, Anthropic can still use your conversations if they are flagged for safety review — a category that is not precisely defined and could be applied broadly. This means the opt-out is not a complete guarantee that your conversations won't be used to train AI. The safety-review exception is determined unilaterally by Anthropic.
What you can do: Go to your account settings and opt out of model training. Understand that flagged conversations may still be used regardless of your opt-out choice.
Anthropic owes you nothing if it suspends your access without warning.
“I.3.b. Anthropic will use reasonable efforts to provide written notice of any Service Suspension to Customer, and resume providing access to the Services, as soon as reasonably possible after the event giving rise to the Service Suspension is cured, where curable. Anthropic will have no liability for any damage, liabilities, losses (including any loss of data or profits), or any other consequences that Customer may incur because of a Service Suspension.”
Quoted verbatim from the agreement What this means: If Anthropic suspends your access to the service — even without warning — it bears zero financial responsibility for any losses you suffer as a result, including lost data or lost business profits. For businesses that rely on the API, this could mean significant uncompensated harm.
What you can do: Build redundancy into any business-critical systems that rely on Anthropic's API, and maintain local backups of important data and outputs.
Feedback and flagged content bypass your training opt-out entirely.
“Our use of Materials. 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, 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 to improve our ability to detect harmful content, enforce our policies, or advance our safety research.”
Quoted verbatim from the agreement What this means: The Consumer Terms of Service confirm the same training policy: your inputs and outputs are used for model training unless you opt out, but the opt-out is bypassed whenever you give feedback or your content is flagged. This is a structural limitation on the opt-out right that most users would not anticipate.
What you can do: Opt out of training in account settings and avoid using feedback buttons on sensitive conversations.
Anthropic trains AI on scraped internet data and purchased datasets.
“Anthropic obtains personal data from third party sources in order to train our models. Specifically, we train our models using data from the following sources: Publicly available information via the Internet Datasets that we obtain through commercial agreements with third party businesses Data that our users or crowd workers provide, including Inputs and Outputs from our Services (unless users opt out) Feedback that users explicitly provide about our Services Materials flagged for safety, security, or policy review Data that we generate internally”
Quoted verbatim from the agreement What this means: Anthropic trains its models on a broad range of data including publicly scraped internet content, commercially purchased datasets, and user conversations. This means personal data about people who have never used Anthropic's services may be included in training data without their knowledge or consent.
What you can do: If you believe your personal data has been included in Anthropic's training data without consent, you can submit a deletion or objection request to [email protected].
Anthropic can share your data with government agencies on its own judgment.
“Pursuant to regulatory or legal requirements, safety, rights of others, and to enforce our rights or our terms. We may share personal data with government authorities, law enforcement, or other third parties where, based on the information available to us, we have a good-faith belief that disclosure is reasonably necessary to (i) comply with applicable law, regulation or legal process, including for legal, tax or accounting purposes, or in response to an enforceable governmental request; (ii) prevent serious harm to any person or to property; (iii) detect, prevent, or otherwise address fraud or other illegal activity; or (iv) enforce our terms, or protect the rights, property, security, or safety of Anthropic, our users, or others.”
Quoted verbatim from the agreement What this means: Anthropic can share your personal data with government authorities and law enforcement based on its own 'good-faith belief' that disclosure is necessary. The standard is self-assessed and broad, covering not just legal compulsion but also Anthropic's own judgment about harm prevention or policy enforcement.
What you can do: Be aware that conversations with Claude may be disclosed to law enforcement without a court order if Anthropic believes it is necessary; avoid sharing sensitive personal information in chats.
Anthropic collects your device location, IP address, and advertising identifiers.
“Device and Connection Information: Consistent with your device or browser permissions, your device or browser automatically sends us information about when and how you install, access, or use our Services. This includes information such as your device type, operating system information, browser information and web page referers, mobile network, connection information, mobile operator or internet service provider (ISP), time zone setting, IP address (including information about the location of the device derived from your IP address), identifiers (including device or advertising identifiers, probabilistic identifiers, and other unique personal or online identifiers), and device location.”
Quoted verbatim from the agreement What this means: Anthropic automatically collects a wide range of device and network data including your IP address, advertising identifiers, probabilistic identifiers, and device location. Probabilistic identifiers can be used to track you across devices and services even without cookies, making this a broad tracking capability.
What you can do: Use a VPN to mask your IP address and location; review your device's advertising ID settings and reset or limit ad tracking where possible.
Anthropic ignores your browser's Do Not Track privacy signal.
“Do Not Track. Some web browsers provide a "Do Not Track" option. We currently do not respond to browser "Do Not Track" signals.”
Quoted verbatim from the agreement What this means: Anthropic explicitly ignores browser Do Not Track signals, meaning your browser's privacy preference is disregarded. This is a deliberate choice not to honor a widely supported user privacy signal.
What you can do: Use browser extensions that block tracking scripts (e.g., uBlock Origin) rather than relying on Do Not Track signals, which Anthropic ignores.
Your data can be transferred to a new owner if Anthropic is acquired.
“As part of a significant corporate event: If Anthropic is involved in a merger, corporate transaction, bankruptcy, or other situation involving the transfer of business assets, Anthropic will receive or disclose your personal data as part of these corporate transactions.”
Quoted verbatim from the agreement What this means: If Anthropic is acquired, merges with another company, or goes bankrupt, your personal data — including all your conversation history — can be transferred to the new owner. You have no right to object to or opt out of this transfer, and the new owner may have different privacy practices.
What you can do: There is no opt-out for this; if Anthropic is acquired, review the acquiring company's privacy policy and consider deleting your account and data beforehand if you are concerned.
Your data transfers to new owners if Anthropic is sold.
“As part of a significant corporate event: If Anthropic is involved in a merger, corporate transaction, bankruptcy, or other situation involving the transfer of business assets, Anthropic will receive or disclose your personal data as part of these corporate transactions.”
Quoted verbatim from the agreement What this means: If Anthropic is acquired, merges with another company, or goes bankrupt, all of your personal data — including your conversation history — can be transferred to the new owner. You have no right to object to or opt out of this transfer, and the new owner may have very different privacy practices.
What you can do: There is no direct action you can take to prevent this. If Anthropic announces a major corporate transaction, consider deleting your account and data before the deal closes.
Anthropic can transfer your contract to anyone without your consent.
“No assignment. These Terms may not be transferred or assigned by you without our prior written consent, but may be assigned by us without restriction.”
Quoted verbatim from the agreement What this means: Anthropic can transfer your contract and all associated rights to any other company without asking your permission. You, however, cannot transfer your account to anyone else without Anthropic's written approval. This is a one-sided arrangement that benefits only Anthropic.
What you can do: There is no action you can take to prevent Anthropic from assigning these Terms. Monitor communications from Anthropic for any notices of assignment.
Deleting your account requires emailing support, not a button.
“You may close your Account at any time by contacting us at [email protected].”
Quoted verbatim from the agreement What this means: Account deletion requires emailing Anthropic's support team — there is no self-service delete button in the account settings. This is a manual process that depends on Anthropic's response time, meaning your data is not immediately deleted when you decide to leave.
What you can do: Email [email protected] to request account closure. Follow up if you do not receive confirmation within a reasonable time.
Cookies track your usage and are used for marketing purposes.
“Cookies & Similar Technologies: We and our service providers use cookies, scripts, or similar technologies (”
Quoted verbatim from the agreement What this means: Cookies) to manage the Services and to collect information about you and your use of the Services. These technologies help us to recognize you, customize or personalize your experience, market additional products or services to you, and analyze the use of our Services to make them safer and more useful to you. For more details about how we use these technologies, and your opt-out controls and other options, please visit our Cookie Policy.
What you can do: Visit Anthropic's Cookie Policy page and use the opt-out controls provided there. You can also manage cookies through your browser settings.
Your employer can read your chats if you use a work email.
“Business Domains. If you use an email address owned by your employer or another organization, your Account may be linked to the organization's enterprise account with us and the organization's administrator may be able to monitor and control the Account, including having access to Materials (defined below). We will provide notice to you before linking your Account to an organization's enterprise account. However, if the organization is responsible for notifying you or has already informed you that it may monitor and control your Account, we may not provide additional notice.”
Quoted verbatim from the agreement What this means: If you sign up with a work email address, your employer may gain access to all your Claude conversations. Anthropic may not warn you about this if your employer has already told you monitoring is possible. This means personal conversations conducted on a work-email account could be visible to your employer.
What you can do: Use a personal email address to create your Claude account if you want to keep your conversations private from your employer.
Anthropic can transfer your account and data to any company without your consent.
“No assignment. These Terms may not be transferred or assigned by you without our prior written consent, but may be assigned by us without restriction.”
Quoted verbatim from the agreement What this means: Anthropic can transfer your account and all associated data to any other company — for example, in a merger or acquisition — without asking your permission. You, however, cannot transfer your account to anyone else without Anthropic's written approval. This is a one-sided arrangement.
What you can do: Be aware that a corporate acquisition of Anthropic could result in your data and account being transferred to a new owner under different management without your consent.
Anthropic can suspend your access with no notice in urgent situations.
“We may modify, suspend, or discontinue the Services or your access to the Services. Where those actions will materially impact your use of the Services, we will endeavor to notify you in accordance with this section. Where feasible, we will provide you with at least 30 days' advance notice of any such changes to the Services that would materially limit or reduce the features, availability or functionality of the Services. However, there may be urgent situations (such as preventing abuse, responding to legal requirements, or addressing security and operability issues) where providing advance notice is not feasible.”
Quoted verbatim from the agreement What this means: Anthropic can suspend or shut down the service with as little as 30 days' notice, and in urgent situations with no notice at all. If you rely on Claude for important workflows, you could lose access suddenly. The 'urgent situations' exception is broadly defined and could cover many scenarios.
What you can do: Avoid building critical business or personal workflows that depend entirely on Claude without a backup plan, as access can be suspended or terminated with little warning.
Everything you type or upload to Claude is collected and stored.
“Inputs and Outputs: You are able to interact with our Services in a variety of formats, including but not limited to chat, coding, agentic sessions (where Claude performs multi-step tasks or takes actions on your behalf), and connected services. The content you submit to the Services through these interactions, including via third-party applications, services, and content you choose to upload, integrate or interact with using our Services, are your "Inputs". Inputs generate responses and actions ("Outputs"). If you include personal data or reference external content in your Inputs, we will collect that information and it may be reproduced in your Outputs.”
Quoted verbatim from the agreement What this means: Everything you type, upload, or connect to Claude — including any personal data you include — is collected and stored. This includes sensitive information you might share in the course of a conversation, such as health details, financial information, or personal circumstances. That data can also appear in your Outputs.
What you can do: Avoid sharing sensitive personal information (health, financial, identifying details) in your Claude conversations unless necessary, and review your privacy settings regularly.
Your conversations are used to train Anthropic's AI models by default.
“Anthropic obtains personal data from third party sources in order to train our models. Specifically, we train our models using data from the following sources: Publicly available information via the Internet Datasets that we obtain through commercial agreements with third party businesses Data that our users or crowd workers provide, including Inputs and Outputs from our Services (unless users opt out) Feedback that users explicitly provide about our Services Materials flagged for safety, security, or policy review Data that we generate internally”
Quoted verbatim from the agreement What this means: Anthropic trains its AI models using your conversations (unless you opt out), publicly scraped internet data, and commercially purchased datasets. Your interactions with Claude directly feed into future model development. The opt-out only partially protects you, as safety-flagged conversations are always used.
What you can do: Opt out of model training in your account settings to reduce (though not eliminate) how your conversations are used for AI training.
Anthropic automatically collects your location, device ID, and ad identifiers.
“Device and Connection Information: Consistent with your device or browser permissions, your device or browser automatically sends us information about when and how you install, access, or use our Services. This includes information such as your device type, operating system information, browser information and web page referers, mobile network, connection information, mobile operator or internet service provider (ISP), time zone setting, IP address (including information about the location of the device derived from your IP address), identifiers (including device or advertising identifiers, probabilistic identifiers, and other unique personal or online identifiers), and device location.”
Quoted verbatim from the agreement What this means: Anthropic automatically collects a wide range of device and network data every time you use the service, including your IP address (used to infer your location), advertising identifiers, and probabilistic identifiers that can track you across sessions. This goes beyond what is strictly necessary to provide the service.
What you can do: Use a VPN to mask your IP address and location, and review your device's advertising ID settings to limit tracking identifiers.
Anthropic uses your company name and logo in marketing by default.
“Anthropic may use Customer's name and logo to publicly identify Customer as a customer of the Services; provided that Customer may opt-out via this request form.”
Quoted verbatim from the agreement What this means: By default, Anthropic can publicly use your organization's name and logo in its marketing materials. You are opted in automatically and must actively submit a form to opt out. This could be embarrassing or commercially sensitive for organizations that prefer not to be publicly associated with AI tools.
What you can do: If you are a commercial or K-12 customer and don't want your name used in Anthropic's marketing, submit the opt-out request form as soon as you sign up.
Cookies are used to market additional products to you.
“Cookies & Similar Technologies: We and our service providers use cookies, scripts, or similar technologies ("Cookies") to manage the Services and to collect information about you and your use of the Services. These technologies help us to recognize you, customize or personalize your experience, market additional products or services to you, and analyze the use of our Services to make them safer and more useful to you.”
Quoted verbatim from the agreement What this means: Anthropic and its service providers use cookies and tracking scripts for marketing purposes, not just for core service functionality. The policy references a separate Cookie Policy for opt-out controls, meaning you need to find and read another document to understand your options.
What you can do: Visit Anthropic's Cookie Policy to find and use the available opt-out controls for non-essential tracking cookies.
Subscription payments are non-refundable unless the law says otherwise.
“Except as expressly provided in these Terms or where required by law, all payments are non-refundable. Please check your order carefully before confirming it”
Quoted verbatim from the agreement What this means: Subscription fees and other payments to Anthropic are non-refundable except where the law requires otherwise. If you pay for a subscription and then decide the service doesn't meet your needs, you generally cannot get your money back.
What you can do: Try any free tier or evaluation period before committing to a paid subscription, and confirm your order details carefully before payment.
Your data is transferred to US servers regardless of where you live.
“Anthropic is a global company, and when you access our Services, your personal data is transferred to our servers in the US, or to other countries outside the European Economic Area ("EEA") and the UK for the purposes outlined in Section 2 of this Privacy Policy, for example to provide you with the Service, or to train our models, or to conduct research.”
Quoted verbatim from the agreement What this means: All users' data, including those in the EU and UK, is transferred to the United States and potentially other countries. While Anthropic states it uses standard contractual clauses and adequacy decisions as safeguards, US-based data is subject to US surveillance laws, which offer fewer protections than EU law.
What you can do: EU/UK users can contact [email protected] or their local data protection authority if they have concerns about international data transfers.
You can delete conversations; back-end removal happens within 30 days.
“Deletion: the right to request that we delete personal data collected from you when you use our Services, subject to certain exceptions. 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.”
Quoted verbatim from the agreement What this means: You can delete individual conversations and they will be removed from your visible history immediately, with back-end deletion within 30 days. A broader account deletion right also exists, though it is subject to exceptions. This is a meaningful user-protective feature.
What you can do: Use the conversation deletion feature regularly for sensitive chats; submit a full data deletion request to [email protected] to remove your account data.
Anthropic does not sell your data and honors global privacy controls.
“Sale & targeted Anthropic marketing of its products and services: Anthropic does not "sell" your personal data as that term is defined by applicable laws and regulations. You can opt-out of sharing your personal data for targeted advertising to promote our products and services, and we will honor global privacy controls.”
Quoted verbatim from the agreement What this means: Anthropic explicitly states it does not sell personal data and allows users to opt out of targeted advertising for Anthropic's own products. It also honors global privacy controls (like the GPC browser signal), which is a user-protective commitment.
What you can do: Enable Global Privacy Control in your browser to automatically signal your opt-out preference to Anthropic.
Anthropic does not make legally significant decisions about you automatically.
“Automated decision-making: Anthropic does not engage in decision making based solely on automated processing or profiling in a manner which produces a legal effect (i.e., impacts your legal rights) or significantly affects you in a similar way (e.g., significantly affects your financial circumstances or ability to access essential goods or services).”
Quoted verbatim from the agreement What this means: Anthropic commits to not making legally significant decisions about you using only automated systems. This means a human must be involved in any decision that could meaningfully affect your rights or access to services.
Anthropic disclaims responsibility for inaccurate AI outputs.
“Reliance on Outputs and Actions. Artificial intelligence and large language models are frontier technologies that are still improving in accuracy, reliability and safety. When you use our Services, you acknowledge and agree: Outputs may not always be accurate and may contain material inaccuracies even if they appear accurate because of their level of detail or specificity. Actions may not be error free or operate as you intended. You should not rely on any Outputs or Actions without independently confirming their accuracy. The Services and any Outputs may not reflect correct, current, or complete information. Outputs may contain content that is inconsistent with Anthropic's views.”
Quoted verbatim from the agreement What this means: By using Claude, you agree that you understand its outputs may be wrong, even when they sound confident and detailed. Anthropic disclaims responsibility for any harm caused by inaccurate AI outputs. This is especially important for medical, legal, financial, or safety-critical questions.
What you can do: Always independently verify important information Claude provides, especially for medical, legal, financial, or safety-related topics. Do not take significant actions based solely on Claude's outputs.
EEA users get a 14-day full-refund cancellation window.
“f. Additional cancellation rights. You have a legal right to change your mind and cancel the Subscription within 14 days of entering into the Subscription without giving a reason. To exercise the right to cancel in the 14-day cancellation period, you must inform us of your decision to cancel the Subscription by making a clear statement to us of such decision before the cancellation period has expired. If you cancel the Subscription under Section 6(f)(1), we will reimburse you all payments received from you for the cancelled Subscription. We will make the reimbursement without undue delay, and not later than 14 days after the day on which we are informed about your decision to cancel the Subscription.”
Quoted verbatim from the agreement What this means: EEA/Switzerland consumers get a 14-day cooling-off period after subscribing, during which they can cancel and receive a full refund with no questions asked. This is a legally required consumer protection right that Anthropic explicitly acknowledges in the Terms.
What you can do: If you subscribe and change your mind, cancel within 14 days via the customer portal or by emailing [email protected] to get a full refund.
Business and school API users get stronger no-training protections.
“As between the parties and to the extent permitted by applicable law, Anthropic agrees that Customer (a) retains all rights to its Inputs, and (b) owns its Outputs. Anthropic disclaims any rights it receives to the Customer Content under these Terms. Subject to Customer's compliance with these Terms, Anthropic hereby assigns to Customer its right, title and interest (if any) in and to Outputs. Anthropic may not train models on Customer Content from Services.”
Quoted verbatim from the agreement What this means: For commercial API customers and K-12 schools, Anthropic explicitly states it will not train its models on their content and assigns full ownership of outputs to the customer. This is a stronger protection than what consumer users of Claude.ai receive, where training opt-out has exceptions.
What you can do: If you are a business or school using the API, your data has stronger protections by default. Consumer users of Claude.ai should manually opt out of training in account settings.
The 14-day refund window is one-time only and does not reset on resubscription.
“14-day cancellation period will not reset if you change subscription tiers or cancel and then resubscribe, as you have already had an opportunity to test the Services.”
Quoted verbatim from the agreement What this means: The 14-day cooling-off period (a legal right in the EU/EEA) is a one-time right. If you cancel and resubscribe, or upgrade/downgrade your plan, you do not get a fresh 14-day window. This is worth knowing if you are considering switching subscription tiers.
What you can do: Use your 14-day cancellation right carefully on your first subscription; it will not be available again if you cancel and rejoin later.
Business API customers get a firm no-training guarantee; consumers do not.
“As between the parties and to the extent permitted by applicable law, Anthropic agrees that Customer (a) retains all rights to its Inputs, and (b) owns its Outputs. Anthropic disclaims any rights it receives to the Customer Content under these Terms. Subject to Customer's compliance with these Terms, Anthropic hereby assigns to Customer its right, title and interest (if any) in and to Outputs. Anthropic may not train models on Customer Content from Services.”
Quoted verbatim from the agreement What this means: For commercial API and K-12 customers, Anthropic explicitly states it will not train its models on their content and assigns full ownership of outputs to the customer. This is a stronger protection than what individual consumer users receive, where training opt-out has exceptions.
What you can do: If data privacy is critical for your use case, consider using Anthropic's commercial API tier rather than the consumer Claude.ai product, as it comes with a no-training guarantee.
Anthropic provides self-service tools to manage your privacy settings.
“Anthropic gives you access to a variety of tools to help you manage your data. You can access these in your Privacy Settings.”
Quoted verbatim from the agreement What this means: Anthropic provides self-service privacy controls through account settings, including the ability to opt out of model training and manage third-party integrations. This is a positive user protection that goes beyond the minimum required in many jurisdictions.
What you can do: Visit your Privacy Settings to review and adjust your data preferences, including opting out of model training.
Anthropic explicitly bans all child exploitation content and reports it to authorities.
“Do Not Compromise Children's Safety This includes using our products or services to: Create, distribute, or promote child sexual abuse material ("CSAM"), including AI-generated CSAM Facilitate the trafficking, sextortion, or any other form of exploitation of a minor Facilitate minor grooming, including generating content designed to impersonate a minor Facilitate child abuse of any form, including instructions for how to conceal abuse Promote or facilitate pedophilic relationships, including via roleplay with the model Fetishize or sexualize minors, including in fictional settings or via roleplay with the model Note: We define a minor or child to be any individual under the age of 18 years old, regardless of jurisdiction. 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.”
Quoted verbatim from the agreement What this means: Anthropic has an explicit, comprehensive policy against any use of Claude to harm children, including AI-generated CSAM, grooming, and exploitation. Notably, they commit to reporting detected CSAM to relevant authorities, which is a strong protective stance that goes beyond what many platforms disclose.
Frequently asked questions
Is Anthropic's Terms & Conditions agreement fair?
- BeforeYouClick grades Anthropic E on an A–E scale, where A is the fairest. Its agreement contains 15 critical clauses, 19 warnings and 11 good practices.
What are the biggest problems with Anthropic's terms?
- The most serious clauses we flagged are: Conversations Used for AI Training by Default, Biometric and Government ID Data Collected and Vague Data Retention With No Fixed Timeframes.
How long does it take to read Anthropic's terms?
- Anthropic's agreement takes about 65 minutes to read in full, across 45 clauses we reviewed. The summary on this page covers the same ground in under a minute.
When was Anthropic's agreement last analyzed?
- We last read Anthropic's Terms & Conditions on August 1, 2026. Companies change their terms without notice, so check the original document for the current version.
Is this analysis legal advice?
- No. This analysis is AI-generated, can be incomplete or wrong, and is not legal advice. Use it as a starting point and read the original agreement before you agree to it.
More agreements to read
- OpenAIAIGrade E
- PosteoProductivityGrade A
- Mullvad VPNSecurityGrade A
- MurenaSecurityGrade B
- TutaProductivityGrade B
- NordVPNSecurityGrade B