What are the intellectual property risks of training a custom AI model on my products?

Quick Answer

The two main risks are input infringement and data leakage. If you train on images you do not own (competitor photos, scraped Pinterest boards, stock you have not licensed), the output is a derivative work that can infringe copyright. The second risk is platform-level: on tools without strict data isolation, your private product imagery could influence outputs for other users.

Risk Checklist

• Input ownership: only train on images you own or have licensed. Your own studio shots and lifecycle photography are safe. Scraped reference boards are not.
• Platform privacy: confirm the tool isolates your uploaded Assets from the public base model and from other tenants.
• Trade secrets: be cautious uploading unreleased prototypes to web-based AI tools until you have verified their data handling.
• Likeness: if your training set includes recognizable people, you need releases, not just copyright clearance.

How Nightjar Handles This

Nightjar isolates user data. When you upload product Assets or build a custom Photography Style from reference imagery, that data drives your Generations only. Your private product photos are not used to train the public base model that other users access.