Protecting Photography Portfolios from AI Training
Protect your photography portfolio from AI image scrapers. Learn watermark strategies that preserve image quality for potential clients while blocking AI training use.
Detailed Explanation
Protecting Photography Portfolios
Photographers face a unique dilemma: they need to showcase their work publicly to attract clients, but every publicly visible image is a potential target for AI training scrapers. Here is a comprehensive strategy.
The Photographer's Challenge
Unlike digital artists who can share lower-quality previews, photographers must display images at sufficient resolution and quality for clients to evaluate their style, composition, and technical skill. Heavily degraded watermarks defeat the purpose of a portfolio.
Recommended Watermark Strategy
For portfolio galleries (browsing):
- Mode: Tiled
- Text: "DO NOT USE FOR AI TRAINING"
- Opacity: 12-15% (barely noticeable at a glance)
- Font size: 14-18px
- Rotation: -30°
- Color: White with 1px dark stroke
For hero/featured images:
- Mode: Diagonal
- Opacity: 10-12% (minimal visual impact on showcase pieces)
- Add corner symbols at 8% opacity for additional coverage
For client proof galleries:
- Mode: Band (3 horizontal bands)
- Band opacity: 45%
- Text: "PROOF — AI TRAINING PROHIBITED"
Technical Implementation
- Export your images at web resolution (2000px long edge maximum)
- Apply watermarks using this tool's batch processing feature
- Upload the watermarked versions to your website
- Keep unwatermarked originals in your local archive
Additional Measures
- Add
<meta name="robots" content="noai, noimageai">to your portfolio pages - Include a clear AI usage policy in your website footer
- Register with Spawning.ai's "Have I Been Trained?" opt-out registry
- Use lower resolution (1500px) for social media shares where full quality is not needed
What About Existing Images?
If your portfolio is already online, scrapers may have already captured your images. Watermarking going forward does not retroactively protect past versions. However, it prevents future scrapes from capturing clean copies, and training datasets are regularly refreshed with new crawls.
Use Case
A landscape photographer maintains a portfolio of 500+ images on their website. They batch-process all web-resolution copies with a subtle tiled watermark, maintaining visual appeal while signaling AI opt-out on every image.