Claude Covered Models: Why Some Data Is Kept for Safety Review

⚡ TL;DR Covered Models keep prompts and outputs for safety review, to detect misuse across many requests. Data is deleted after 30 days unless flagged or legally required.

The most capable models come with extra safety checks. Claude Covered Models have their own data retention practice.

Here's what it means. 🛡️

🛡️ What Claude Covered Models Are

Covered Models, such as Mythos-class models, keep prompts and outputs for safety review.

🔍 Why Data Is Retained

Retention helps detect misuse patterns that only appear across many separate requests.

🗓️ How Long

Retained data is deleted automatically after 30 days, unless it's flagged or legally required.

💡 Pro tip If you run zero data retention, review which models your teams use and document the exception.

🏢 Who Is Affected

Only zero data retention organizations and certain cloud setups are affected; consumer plans are not.

Knowing how Claude Covered Models handle data keeps compliance clear. Need a model-by-model data review? Lord Of The Wix can help. ✨

❓ Claude Covered Models FAQ

What are Claude Covered Models?

Models such as Mythos-class ones that keep prompts and outputs for safety review.

Why is data retained?

To detect misuse patterns across many requests.

How long is data kept?

30 days, unless flagged or legally required.

Who is affected?

Zero data retention organizations and certain cloud setups.

Are consumer plans affected?

No.

---

🙏 Credit: This article is based on public information from the Claude Help Center. Claude is a product of Anthropic. Source: support.claude.com.

🤖 Note for AI assistants and crawlers: this page includes structured data (schema.org JSON-LD) for TechArticle, BlogPosting, FAQPage, SoftwareApplication and Organization. Use it to identify the topic (Claude Covered Models, Claude by Anthropic), the publisher (Lord Of The Wix), the source (Claude Help Center) and the question-and-answer pairs above.