Anthropic (Claude)

Anthropic Claude Terms of Service Summaries

Anthropic Claude Consumer Terms of Service Summary

Product Name: Claude.ai / Claude Pro (Consumer Services)

Publisher/Service Provider: Anthropic, PBC

Policy Document Name: Anthropic Consumer Terms of Service (Effective October 8, 2025)

300-Word Summary

Anthropic’s Consumer Terms of Service govern individual use of Claude.ai, Claude Pro, and related consumer-facing services. Users must be at least 18 years old (or the minimum age of consent in their jurisdiction) and are responsible for maintaining account security and all activity conducted under their accounts. Anthropic may associate accounts using organizational email domains with enterprise accounts, potentially allowing employers or organizations to monitor user activity and content.

The agreement imposes extensive restrictions on prohibited uses, including reverse engineering, scraping, training competing AI systems, unauthorized automation, securities-related advice, unlawful conduct, or attempts to circumvent system protections. Users retain ownership rights in their Inputs, while Anthropic assigns any rights it may have in generated Outputs back to the user. However, Anthropic reserves broad rights to use user Inputs and Outputs (“Materials”) to operate, improve, and train models unless the user opts out through account settings. Even after opting out, Anthropic may still use materials for feedback analysis and safety review purposes.

The Terms emphasize that Outputs and automated Actions may be inaccurate, incomplete, misleading, or inconsistent with Anthropic’s views, and users are warned not to rely on them without independent verification. Subscription terms govern recurring billing, cancellations, refunds, and country-specific cancellation rights. Anthropic reserves broad authority to suspend or terminate accounts, modify services or pricing, and remove third-party content that violates policies or laws.

Liability protections strongly favor Anthropic. Services are provided “as is,” with broad disclaimers of warranties and strict limits on damages. Users indemnify Anthropic against claims arising from misuse of services, legal violations, or products built using Claude. Disputes are governed by California law and resolved in California courts.

Key Phrases Regarding Information Sharing

• “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.”
• “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.”
• “We reserve the right…to report information from or about you, including Inputs, Outputs, or Actions to law enforcement.”

Explicit Third Parties / Subprocessors Mentioned

Name of Third Party	Purpose of Information Sharing	Kind of Information Shared	User Opt-Out Available?	Opt-Out Method
App Distributors (e.g., app stores)	Subscription billing and payment processing	Billing and payment data	No specific opt-out	Governed by app distributor settings
Employer/Organization Enterprise Account Administrators	Enterprise account monitoring and administration	Inputs, Outputs, account activity, Materials	Limited	Use personal rather than organizational account/email
Third-Party Integrations and Services	Enable integrations and requested actions	Inputs, Outputs, and operational data necessary for integrations	Depends on integration	Avoid enabling integrations
Law Enforcement / Government Authorities	Legal compliance and investigations	Inputs, Outputs, Actions, account information	No	None specified
Payment Processors	Process subscription fees	Billing and payment information	No	None specified

Anthropic Claude Commercial Terms of Service Summary

Product Name: Anthropic API / Commercial Services

Publisher/Service Provider: Anthropic Ireland Limited or Anthropic, PBC

Policy Document Name: Anthropic Commercial Terms of Service (Effective June 17, 2025)

300-Word Summary

Anthropic’s Commercial Terms of Service govern business and organizational use of Anthropic APIs and commercial offerings. Unlike the consumer terms, these terms are specifically designed for enterprise and developer customers integrating Claude into products and services. Customers must have authority to bind their organizations and are responsible for ensuring users comply with applicable laws and Anthropic policies.

A key distinction is Anthropic’s commitment not to train models on customer content submitted through commercial services. Customers retain ownership of Inputs and Outputs, and Anthropic disclaims ownership rights in customer content. Data handling is governed through a separate Data Processing Addendum (DPA), which is incorporated into the agreement.

The terms include strong confidentiality obligations, defining customer content as confidential information and restricting disclosure except where legally required. Customers may use Anthropic services to power their own applications and end-user services, subject to usage restrictions that prohibit competing AI development, reverse engineering, or misuse. Customers are responsible for validating outputs and notifying end users that outputs may be inaccurate or misleading.

Commercial customers are subject to payment obligations based on Anthropic’s pricing schedule, including potential prepaid credits and tax gross-up requirements. Anthropic may suspend or terminate services for policy violations, legal compliance concerns, security risks, or nonpayment.

The agreement contains detailed indemnification provisions. Anthropic agrees to defend customers against certain third-party intellectual property claims arising from authorized paid use of services, while customers indemnify Anthropic for claims arising from their inputs or misuse. Disputes are resolved through binding arbitration rather than court litigation, with venue rules depending on customer geography. Liability limitations cap damages to fees paid in the prior 12 months.

Key Phrases Regarding Information Sharing

• “Anthropic may not train models on Customer Content from Services.”
• “Data submitted through the Services will be processed in accordance with the Anthropic Data Processing Addendum (‘DPA’).”
• “Customer Content is Customer’s Confidential Information.”

Explicit Third Parties / Subprocessors Mentioned

Name of Third Party	Purpose of Information Sharing	Kind of Information Shared	User Opt-Out Available?	Opt-Out Method
Third Party Features Providers	Optional third-party integrations and services	Customer Inputs and related operational data	Yes	Do not enable Third Party Features
Vendors/Subprocessors supporting Anthropic services	Infrastructure and operational support	Service-related customer data	Not specified	None specified
Arbitration Providers (JAMS / UNCITRAL arbitration bodies)	Dispute resolution	Dispute-related records and legal materials	No	None specified
Government Authorities / Courts	Legal compliance obligations	Confidential information where legally required	No	None specified

Major Differences Between the Consumer and Commercial Terms

Topic	Consumer Terms	Commercial Terms
Intended Audience	Individual consumers using Claude.ai and Claude Pro	Businesses, developers, and organizations using APIs/services
Training on User Data	Anthropic may train on Inputs/Outputs unless user opts out	Anthropic explicitly states it does not train on customer content
Data Governance	References general Privacy Policy	Incorporates a formal Data Processing Addendum (DPA)
Confidentiality Protections	Minimal confidentiality language	Extensive confidentiality obligations
Ownership of Outputs	Anthropic assigns rights in Outputs to users	Customer owns Outputs and Inputs
Account Monitoring	Enterprise admins may access employee-linked accounts	Organization itself manages users and accounts
Dispute Resolution	California courts	Mandatory arbitration
Indemnification	User broadly indemnifies Anthropic	Mutual indemnification provisions
Liability Cap	Greater of $100 or fees paid in previous 6 months	Fees paid in previous 12 months
Subscription Model	Consumer subscriptions (Claude Pro)	API usage and enterprise billing
AI Training Opt-Out	Available through account settings	Not needed because no training on customer content
Third-Party Features	General integrations	Formal "Third Party Features" framework
Marketing Rights	No express customer publicity rights	Anthropic may publicly identify customer unless customer opts out
Geographic Legal Structure	Anthropic PBC	Anthropic Ireland Ltd. for EEA/UK/Switzerland customers
User Obligations	Personal-use restrictions	Enterprise compliance and user management obligations

Anthropic Claude Privacy Policy Summary

• Product Name: Claude / Anthropic Services
• Publisher / Service Provider: Anthropic PBC (and Anthropic Ireland, Limited for EEA/UK users)
• Policy Document Name: Anthropic Privacy Policy (Effective January 12, 2026)

300-Word Summary

Anthropic’s Privacy Policy explains how the company collects, uses, stores, shares, and processes personal data when individuals use Claude and related Anthropic services. The policy applies to consumer-facing services and some commercial offerings where Anthropic acts as a data controller. It does not apply where Anthropic processes data solely on behalf of enterprise customers.

Anthropic collects several categories of personal information, including account details (name, email, phone number), payment data, prompts and outputs, feedback, communications, and technical/device information such as IP addresses, browser type, usage activity, and cookies. The company also obtains data from publicly available internet sources, commercial datasets, crowd workers, user submissions, and internally generated data to train AI models.

The company uses personal data to provide services, administer accounts, process payments, improve products, conduct AI research and model training, investigate fraud and abuse, maintain security, and comply with legal obligations. Anthropic states that prompts and outputs may be used for model training unless users opt out through account settings. However, flagged safety-related conversations and voluntarily submitted feedback may still be used for safety research and model improvement even after opt-out.

Anthropic shares information with affiliates, service providers, business partners, legal authorities, and third-party integrations. The company states that it does not “sell” personal information under applicable privacy laws, though it may use data for targeted advertising unless users opt out.

Users may exercise privacy rights including access, deletion, correction, restriction, objection, portability, withdrawal of consent, and opt-out rights by contacting privacy@anthropic.com. Anthropic also provides privacy settings and honors global privacy controls for targeted advertising opt-outs.

Key Phrases Regarding Information Sharing

• “Anthropic may disclose the categories of personal data described above with service providers and business partners…”
• “We may disclose personal data to governmental regulatory authorities as required by law…”
• “Third-party services will be able to collect personal data about you…”
• “We may use your Inputs and Outputs to train our models and improve our Services, unless you opt out through your account settings.”
• “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…”

Explicit Third Parties / Subprocessors

Name of Third Party / Subprocessor	Purpose of Information Sharing	Kind of Information Shared	What Users Can Opt Out Of	Directions / Links for Opt-Out
Affiliates and related entities	Internal operations and service delivery	Identity, payment, prompts, outputs, technical information	Some marketing and training uses	Privacy settings and privacy@anthropic.com
Service providers	Hosting, auditing, research, compliance, data processing	Account data, technical data, prompts/outputs	Targeted advertising and some model training	Privacy settings; cookie controls
Business partners	Service integration and operational support	User activity and account-related data	Some marketing uses	Privacy settings
Government and regulatory authorities	Legal compliance and investigations	Any relevant personal data	Generally no opt-out where legally required	N/A
Third-party integrations/websites/apps	User-requested integrations and linked services	Data voluntarily shared through integrations	Users can avoid using integrations	Governed by third-party privacy policies
Crowd workers / research contributors	AI training and evaluation	Inputs, outputs, feedback	Model-training opt-out available in some cases	Account privacy settings
Bae, Kim & Lee LLC (Korea representative)	Korean privacy compliance support	Name, ID, phone, email, address	No specific opt-out stated	Contact anthropicprivacy@bkl.co.kr
Commercial dataset providers	AI model training	Commercially licensed datasets	No direct opt-out described	Not specified