Anthropic
Claude
AI safety lab. Claude family of models.
Anthropic vs OpenAI vs Google Vertex AI / Gemini on three load-bearing dimensions.
| Field | Anthropic | OpenAI | Google Vertex AI / Gemini |
|---|---|---|---|
| No-train default | Anthropic may not train models on Customer Content (Inputs + Outputs) under Commercial Terms | API and Enterprise: not used for training by default. Consumer ChatGPT: opt-out available. | Vertex AI: customer data not used to train base models. Consumer Gemini: distinct policy. |
| IP indemnity | Pending verification (Claude Enterprise terms) | Copyright Shield (ChatGPT Enterprise, API) | Indemnified Generative AI Services (capped, named services) |
| Retention default | API: 30 days for safety. Enterprise: 0 days available. | API: 30 days default. Zero-data-retention available for Enterprise. | Vertex 30 days default; configurable lower |
Usage Policy (effective September 15, 2025)
Our Usage Policy (also referred to as our "Acceptable Use Policy" or "AUP") applies to anyone who can submit inputs to Anthropic's products and/or services, including via any authorized resellers or passthrough access, all of whom we refer to as "users."
Anthropic may not train models on Customer Content (Inputs + Outputs) under Commercial Terms
Anthropic may not train models on Customer Content from Services. "Inputs" means submissions to the Services by Customer or its Users and "Outputs" means responses generated by the Services to Inputs (Inputs and Outputs together are "Customer Content").
Customer retains all rights to Inputs; owns Outputs
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.
Pending verification (Claude Enterprise terms)
API: 30 days for safety. Enterprise: 0 days available.
At least 60 days notice before retirement for publicly released models
Anthropic notifies customers with active deployments for models with upcoming retirements, providing at least 60 days notice before model retirement for publicly released models.
Responsible Scaling Policy v3.3 (effective May 26, 2026)
Anthropic's Responsible Scaling Policy. Last updated May 26, 2026. Version 3.3 (effective May 26, 2026): revises threshold for novel chemical/biological weapons production, refines approach to off-cycle updates regarding the risks of individual models, and makes minor terminology changes.
Public AI strategic partnerships tracked by Anteroom that involve Anthropic. Each row links to the primary-source-anchored deal page with structural read and pattern instances.
- Google ↔ Anthropicstrategic_investactiveannounced 2025-10-23
Anthropic expands Google Cloud TPU usage (up to 1M TPUs, >1 GW online in 2026, tens of billions)
- Anthropic ↔ Salesforceisv_embeddedactiveannounced 2025-10-14
Claude becomes preferred Agentforce model for regulated industries with deep Slack integration
- Deloitte ↔ Anthropicpartner_programactiveannounced 2025-10-06
Deloitte deploys Claude to 470,000 employees globally, Anthropic largest enterprise deal
- Anthropic ↔ Databricksmodel_hostactiveannounced 2025-03-26
Five-year deal: Claude natively available on Databricks Data Intelligence Platform
- Anthropic ↔ Amazon Web Servicesstrategic_investactiveannounced 2024-11-22
Anthropic-AWS expanded partnership ($8B total investment, AWS as primary cloud and training partner)
- Anthropic ↔ Snowflakemodel_hostactiveannounced 2024-11-20
Snowflake and Anthropic strategic partnership bringing Claude to Cortex AI
- Zoom Communications ↔ Anthropicstrategic_investactiveannounced 2023-05-16
Zoom invests in Anthropic and integrates Claude across Zoom platform starting with Contact Center
No structural changes recorded yet for this vendor. When a change in any of the 7 commitments is detected, a row appears here with the old value, the new value, the effective date, and the primary source.
Not legal advice. Not a substitute for negotiating the actual MSA. The Commitments tracker quotes from public-source vendor documents and links to the original; counsel reviewing a deal still has to read the contract in front of them.