# Content Credentials Basics Reference

For the method in
[Content Credentials and Watermarks: What the Little Badge Actually Means](/articles/content-credentials-watermark-basics).
A quick reference for reading a badge correctly - not a truth verdict.

## Not for

Deciding a business's publishing policy - see
[provenance, watermarking, and Content Credentials](/articles/provenance-and-watermarking) for that.

## What the badge DOES tell you

- [ ] This exact file is bound to a manifest
- [ ] The manifest is signed by an identity the platform trusts (if validated)
- [ ] The manifest declares a specific creation/editing history

## What the badge does NOT tell you

- [ ] Whether the depicted event is true
- [ ] Whether the depicted person consented
- [ ] Whether the caption/context is accurate
- [ ] Whether the signer is honest
- [ ] Whether a file WITHOUT a badge is fake (it usually just means no supporting tool was used)

## What breaks a credential (without making the file fake)

| Action | Typical effect on credential |
| --- | --- |
| Screenshot / screen recording | Broken - new unbound file |
| Re-upload through unaware app | May be stripped |
| Heavy compression / format change | May break |
| Forwarded through messaging apps | Often stripped |

## Reading order when you see a badge

1. Is it actually validated (not just present)?
2. What does it assert - full generation, minor edits, something else?
3. Does that change what you assumed you were looking at?
4. If no badge: move to source-checking methods instead of assuming anything from the absence

## Quick decision

☐ Badge present and validated - note what it actually asserts, still verify context separately
☐ No badge - treat as neutral, use source-check methods (viral claim trace, image limits checklist)