Remove ChatGPT Image Metadata
Every image generated by ChatGPT uses DALL-E 3 and contains hidden C2PA Content Credentials. These credentials trigger "Made with AI" labels on Instagram, TikTok, and Pinterest automatically. StripShot removes them at the binary level with zero quality loss.
What ChatGPT and DALL-E embed in your images
ChatGPT uses DALL-E 3 as its image generation model. DALL-E 3 embeds multiple metadata layers in every output. Each layer is read by different platforms and detection systems.
C2PA Content Credentials (JUMBF)
This is the primary trigger for platform AI labels. DALL-E 3 embeds a C2PA manifest using the JUMBF format. In JPEG files, this data lives in APP11 markers. In PNG files, it uses the caBX chunk.
The manifest contains a cryptographic assertion signed by OpenAI, a creation timestamp, and OpenAI tool identification. Instagram, TikTok, and Pinterest read this manifest and apply AI labels. The signature chain traces back to OpenAI's certificate authority.
XMP DigitalSourceType
DALL-E sets the XMP DigitalSourceType field to trainedAlgorithmicMedia. This is stored in a separate APP1 XMP marker distinct from the EXIF APP1 block.
Some detection systems scan the XMP DigitalSourceType field independently from C2PA. StripShot removes the entire XMP block, not just the DigitalSourceType field.
EXIF Software field
The EXIF Software tag references OpenAI and DALL-E. This is stored in the APP1 EXIF block. It is less commonly checked than C2PA, but some platforms and third-party detection tools scan it. StripShot removes all APP1 EXIF markers entirely.
Note: ChatGPT does not use SynthID
SynthID is a pixel-level watermark developed by Google DeepMind. It is used in Google Gemini images. ChatGPT and DALL-E do not use SynthID. DALL-E relies exclusively on C2PA metadata for provenance. StripShot removes DALL-E C2PA completely.
How to remove ChatGPT image metadata
- Step 1: Generate your image in ChatGPT and download it (JPEG or PNG)
- Step 2: Go to stripshot.app and upload the file
- Step 3: StripShot detects the C2PA manifest in the APP11 marker (JPEG) or caBX chunk (PNG)
- Step 4: Click Strip to remove C2PA, XMP DigitalSourceType, and all EXIF at binary level
- Step 5: Download the clean file and upload to Instagram, TikTok, Pinterest, or LinkedIn
All processing happens in your browser. Files never leave your device. Zero quality loss.
Platform detection by platform
| Platform | Detection method | Result |
|---|---|---|
| C2PA Content Credentials | "Made with AI" label added to post | |
| TikTok | C2PA Content Credentials | AI label applied to post |
| C2PA + perceptual hash | Search suppression of AI-flagged pins | |
| C2PA + EXIF scan | AI disclosure label on image posts | |
| C2PA Content Credentials | "Made with AI" label added to post |
ChatGPT image editing also adds metadata
ChatGPT's image editing features (inpainting, style changes, extending) also embed C2PA credentials in the output. A real photograph edited through ChatGPT will receive AI metadata and can trigger platform labels, even if the original photo had none.
The GPT-4o native image output format (introduced April 2025) embeds the same C2PA Content Credentials as earlier DALL-E 3 output. The metadata format is identical regardless of which ChatGPT model version generated the image.
Frequently asked questions
Does ChatGPT watermark images?
ChatGPT does not add a visible watermark but embeds C2PA Content Credentials in every DALL-E generated image. This metadata identifies the file as AI-generated by OpenAI and triggers automatic Made with AI labels on Instagram, TikTok, and Pinterest. It is invisible but machine-readable.
How do I remove ChatGPT metadata?
Download your ChatGPT image. Upload it to StripShot at stripshot.app. The tool removes the C2PA manifest from the APP11 JPEG marker or caBX PNG chunk, clears the XMP DigitalSourceType field, and strips all EXIF data. Zero quality loss. Files never leave your browser.
What is C2PA and why does ChatGPT use it?
C2PA (Coalition for Content Provenance and Authenticity) is an open standard for embedding signed provenance data in media files. OpenAI embeds C2PA in DALL-E images to identify them as AI-generated. Instagram, TikTok, and Pinterest read this data to apply AI content labels.
Does StripShot remove SynthID from ChatGPT images?
ChatGPT and DALL-E do not use SynthID. SynthID is a Google DeepMind technology used in Google Gemini images. DALL-E uses C2PA metadata. StripShot removes DALL-E C2PA completely.
Will my ChatGPT image be flagged after stripping?
No. Instagram, TikTok, and Pinterest use C2PA Content Credentials as the primary trigger for AI labels. After stripping with StripShot, the file contains no C2PA data and the label will not appear.
Is removing ChatGPT metadata free?
Yes. StripShot is free for 3 images per day with no account required. Pro is $9/month for unlimited images and additional features.
Related guides
Remove DALL-E Metadata
Full technical breakdown of DALL-E metadata layers and how to strip them.
ChatGPT + DALL-E Metadata Guide
Deep dive on C2PA in ChatGPT images and why it triggers AI labels.
Bypass AI Detection on Instagram
Full guide for preventing the Made with AI label on Instagram posts.
What Are C2PA Content Credentials
How the C2PA standard works and why platforms rely on it for AI detection.
Pricing
Clean 3 images free. Then go unlimited.
No card needed to start. Upgrade anytime.
- 2 images/day free
- 1 video/day free
- AI detection (70+ signatures)
- 100% private, no uploads
- Unlimited image cleanings
- 10 videos/day included
- Batch processing
- Smart Strip & hex viewer
- Edit & write EXIF tags
- Unlimited image cleanings
- Unlimited video cleanings
- Strip Sora, Runway, Kling
- MP4 + MOV support
- New generators added regularly