This analysis describes what Anthropic's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
The provision establishes a dual-consent framework where users may restrict training use through opt-out but retain defined carve-outs that permit training use for safety-related flagging and user-initiated reporting. This structure maintains model improvement capabilities for specified institutional purposes while providing users discretionary control over routine training use.
Users may prevent their conversation data from being used for model training by configuring account settings, but this election does not apply to conversations identified through safety review processes or materials the user reports directly to Anthropic. The terms specify that training use continues for these defined categories regardless of opt-out status.
How other platforms handle this
If you are in the 'Designated Countries', LinkedIn Ireland Unlimited Company ('LinkedIn Ireland') will be the controller of your personal data provided to, or collected by or for, or processed in connection with our Services. If you are outside of the Designated Countries, LinkedIn Corporation will ...
Crusoe (Sees code data for inference): We manage Crusoe's compute for training some of our custom models, as well as hosting some of our custom models. Modal (Sees code data for inference): We manage Modal's compute for training some of our custom models, as well as hosting some of our custom models...
We use information about you to provide, improve, and develop our products and services, personalize your experience, show you relevant content and ads, and communicate with you. We draw inferences about your interests and preferences based on your activity on Pinterest and elsewhere.
Monitoring
Anthropic has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 10 platforms.
"We may use your Inputs and Outputs to train our models and improve our Services, unless you opt out through your account settings. Even if you opt-out, we will use Inputs and Outputs for model improvement when: (1) your conversations are flagged for safety review to improve our ability to detect harmful content, enforce our policies, or advance AI safety research, or (2) you've explicitly reported the materials to us (for example via our feedback mechanisms).— Excerpt from Anthropic's Anthropic Privacy Policy
We read the privacy policies and terms of service of 38 AI platforms. Here is what they say about training, retention, arbitration, and liability.
Professional Governance Intelligence
Need to monitor specific governance provisions?
Professional includes provision-level monitoring, governance timelines, regulatory mapping, and audit-ready analysis.
Built from archived source documents, structured governance mappings, and historical version tracking.
The provision establishes a dual-consent framework where users may restrict training use through opt-out but retain defined carve-outs that permit training use for safety-related flagging and user-initiated reporting. This structure maintains model improvement capabilities for specified institutional purposes while providing users discretionary control over routine training use.
Users may prevent their conversation data from being used for model training by configuring account settings, but this election does not apply to conversations identified through safety review processes or materials the user reports directly to Anthropic. The terms specify that training use continues for these defined categories regardless of opt-out status.
ConductAtlas has identified this type of provision across 1 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Anthropic.