Hello,
I’m building a commercial service (French e-commerce, personal project)
that transforms children’s photos into stylized superhero-themed
illustrations (comic book / pop art / cartoon styles), delivered as
printed framed artwork. Parents upload a photo of their own child
(with their explicit consent) and receive a stylized, non-photorealistic
illustration as a physical product.
I’ve run a small test batch (24 calls, gemini-3-pro-image model,
paid tier) using photos of my own children, with prompts requesting
comic book / pop art / illustrated superhero transformations, and
all 24 succeeded without any safety block.
Before building a production service around this, I’d like to confirm:
-
Is this use case (transforming a real child’s photo into a
stylized, non-sexual, non-realistic illustration, for a
commercial print product, with parental consent) considered
an acceptable use under the Generative AI Prohibited Use Policy? -
Are there specific safety thresholds or restrictions I should be
aware of at scale (e.g., photorealistic outputs vs. stylized/
illustrated outputs) that aren’t apparent from small-scale testing? -
Is there a way to get a documented confirmation of this use case’s
compliance before scaling to production?
Thank you for any guidance.