Compare privacy rights governance provisions between OpenAI and Anthropic. Provisions are extracted from monitored governance documents and classified by severity.
This provision establishes OpenAI's default authorization to use submitted Content for service improvement and product development, while creating carve-outs for enterprise and API customers. The opt-out mechanism allows individual users to restrict use of their Content for model training purposes, affecting how user-generated data flows into OpenAI's product development processes.
Consumer impact
Users operate under a default authorization permitting OpenAI to use their Content for service maintenance and new product development. Users who wish to restrict use of their Content specifically for model training may opt out through designated mechanisms, but the terms authorize continued use for other specified purposes unless the user exercises the opt-out option.
Opt-out available
No opt-out available
Actual clause text
We may use Content to provide, maintain, and improve our Services and to develop new products and services. We won't use Content from ChatGPT Team, ChatGPT Enterprise, or our API (outside of our models) to train our models. If you don't want us to use your Content to train our models, you can opt out by following the instructions in our Help Center or by visiting our privacy portal.
AI-extracted from source document. Verify against original for legal use.
The provision establishes a tiered consent model where the opt-out right does not apply uniformly—certain categories of user-submitted content (feedback and flagged materials) remain available for model training regardless of opt-out status. This creates an operational distinction between materials subject to user preference settings and materials subject to mandatory training use.
Consumer impact
Users can prevent use of their submitted materials for model training through account settings, but this opt-out does not apply to feedback they provide or to materials Anthropic flags internally for safety purposes. Materials falling within these two exceptions will be used for model training and safety research independent of the user's opt-out election.
Opt-out available
No opt-out available
Actual clause text
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.
AI-extracted from source document. Verify against original for legal use.
The Department of Defense designated Anthropic a supply chain risk after the company refused to rem…
AI Difference AnalysisProfessional
Stripe's arbitration clause is narrower than Amazon's in one key respect: it includes a small claims court carve-out that Amazon's clause does not. PayPal's clause is the most aggressive of the three, explicitly waiving jury trial rights in addition to class action rights. From a compliance perspective, Amazon presents the lowest risk for B2B contracts while PayPal creates the highest exposure for consumer-facing applications subject to CFPB oversight.