Anthropic will let companies keep AI activity data in their own cloud

Enterprise Frontier Safeguards pairs zero data retention with automated misuse detection, and stores the data in your cloud account instead of Anthropic's.

Anthropic announced Enterprise Frontier Safeguards on 1 September. It combines what the company calls "the privacy of zero data retention (ZDR) with state-of-the-art safeguards for detecting misuse". The point is that you can have both, rather than trading one away for the other.

Activity data sits in your own cloud account instead of Anthropic's: Amazon S3, Azure Blob Storage or Google Cloud Storage, and Anthropic says it stays there "under their own encryption keys, access policies, and audit logging". Automated systems scan a rolling window of traffic for signs of serious misuse, such as attempts to build offensive cyber or biological capabilities, or stolen and leaked credentials. Anything flagged goes straight to the customer, and "no human review by Anthropic employees is required". It covers Claude Code, Claude Enterprise, the Claude Platform, Amazon Bedrock, Google's Agent Platform and Microsoft Foundry.

For a midsize business, this is aimed squarely at the objection that stops most AI projects: nobody wants to explain to a client or an auditor that their data went to a third party and stayed there. Holding the keys yourself changes that conversation. It is the same instinct behind keeping your files in your own systems rather than someone else's.

Two caveats before you plan around it. Anthropic says EFS "will be rolling out to customers in phases, starting later this fall", so it is not something you can switch on today, and eligible customers get zero data retention on Fable 5 and 5.1 in the meantime. Anthropic does not charge for EFS, but your cloud provider bills you for the storage and data operations, so the cost moves rather than disappears.

If data handling is the reason AI has stayed on your someday list, this is worth reading before your next vendor review.