How AI is used, and how I disclose it

The model produces the images. I decide what gets made, I sort what reads as the person from what does not, and what does not read does not ship. The model executes; the judgment is human, and it is mine.

Every file I deliver is disclosed as AI-produced. I do not present these portraits as camera photographs, and I do not sell the idea that you cannot tell. What you are buying is directed AI, stated plainly, not a claim that it passes for something else.

The model provider also writes an invisible provenance marker into the pixels at generation. It is theirs, not mine, and it is read by their verification tools rather than by a viewer. I mention it because it is real, and I do not claim any marking beyond it.

I run the AI production from the EU, so the EU transparency rules that took effect on 2 August 2026 apply to my side of the work regardless of where your company sits. Every project is produced to that standard by default, which is the stricter of the two regimes I work across.

US rules point elsewhere. New York's synthetic performer disclosure law, in effect since 9 June 2026, expressly leaves out likenesses of real, identifiable people, and California's transparency act binds the model providers rather than the studios using them. What does bind a US buyer is consent, which is why written, per-person consent is the part I hold firmest. This describes how I work, not a legal opinion on your own publishing.

Production uses a third-party model that processes outside the EU, under your prior consent. Every service that touches your data is listed on the subprocessors page, and how your data is handled is set out in my privacy and data policy.

Corporate buyers are right to ask how their images were made. I would rather answer that up front than have it surface later. Disclosure is the position here, not a footnote.