FAQ

It already has, and it is manageable. Article 50 has applied since 2 August 2026. It does not put a label on every AI image, only on what the Act calls a deep fake, and that definition is wider than most people expect: it covers a photorealistic model who does not exist, and a product rendered to look better than the real one. It does not cover colour correction, a real pack on an AI background, or rearranging products you already photographed. AICO is the deployer under the Act, so the visible disclosure is our job rather than yours. We tell you which assets need a mark, design one that does not wreck the art direction, and apply it before delivery. Separately, Article 4 asks every company using AI to take measures to support its team's AI literacy. There is no exam and no official certificate: you train your people and keep a record, and our training module gives you both. The heavier high-risk rules do not apply to normal beauty marketing and have moved to December 2027.

No certificate is required by the law, that is not how the Act works. You need the training done and a record kept, and we can give you both. Ask us about the AICO EU AI Act training module, a ready-made narrated course your team runs in under an hour, with a completion record you keep on file.

Not for typical beauty content. Some platforms may add their own markers automatically.

We track EU, UK, US and retailer policy developments continuously.

Pricing depends on scope and volume. We offer project pricing and monthly retainers starting from €500 a month.

Yes, production is faster and overall costs are lower.

Your usage rights are included in your agreed project fee.

We offer discounted rates for multiple products or complete ranges.

Packshots, packaging files, shade references, brand guidelines and any inspiration. For larger collections, product samples help with colour accuracy.

Yes, we can rebuild or enhance visuals based on your photos.

Most projects deliver within 3 to 7 days, depending on volume.

JPG, PNG, PSD or any retailer-required format.

We generate unique synthetic models that are not based on real individuals.

Yes, we represent all skin tones, undertones and hair types.

Yes, we follow responsible beauty guidelines to keep visuals believable and natural.

We use calibrated workflows, brand references and product samples where possible.

Yes. We lock lighting, angles and styling to ensure visual consistency.

Yes, we support standards for Sephora, Amazon, Douglas, and other major retailers.

Yes. You can request any target audience or beauty profile.

Yes, standard product and lifestyle images are currently accepted across major beauty platforms.

Yes, since 2 August 2026. Standard advertising law still applies, and on top of it the EU AI Act requires a visible disclosure on AI content that counts as a deep fake, which includes most campaign imagery built around a synthetic model. The Act's exemption for artistic and creative work expressly does not cover advertising. We handle the mark on anything we deliver.

Some add invisible metadata at the point of generation. That is the model provider's duty and it does not discharge yours: the Act expects a disclosure a person can actually see, so invisible marking on its own is not enough.

Yes, we provide high-resolution files suitable for print.

Yes, for your own brand or with permission. Trademarks remain owned by the brand.

Yes, as long as they are clearly labelled as concepts and not implied partnerships.

Concepts show creative possibilities. Client work is produced under an approved commercial agreement.

Once delivered, you receive full commercial usage rights exactly as defined in your agreement.

Yes, social media, e-commerce, ads, print, retail and more, unless your agreement specifies otherwise.

Yes. Cropping, resizing or adjusting visuals is permitted within your usage license.

Sometimes, and we do it for you. Since 2 August 2026 the EU AI Act requires a visible disclosure wherever an asset is a deep fake, which in practice means a photorealistic model who does not exist, or a product rendered to look better than the real one. A straight product shot on an AI background does not need one. Outside the EU, UK and US rules still turn mainly on whether imagery misleads. Everything we deliver arrives already marked where marking is due.

Yes, provided they are not based on a real person and the imagery does not suggest a result the product cannot deliver. One thing worth knowing: under the EU AI Act a photorealistic model who does not exist still counts as a deep fake, so the asset carries a visible AI mark. We apply it.

Yes. Retailer policy is separate from the law and it is moving: Amazon now asks sellers to tag images containing photorealistic AI-generated people. Anything we deliver already carries the disclosure the EU AI Act requires, so it is ready for the strictest of them. We check platform rules per launch, because they change.

Disclosure is required rather than merely recommended wherever an asset is a deep fake under the Act: a photorealistic model who does not exist, or a product rendered to look better than the real one. It is not required for colour correction, a real pack on an AI background, or a rearrangement of products you already shot. We mark what needs marking and tell you why.

Yes, AI can generate synthetic models, and no model release is required if the person is not based on a real identifiable individual.

General aesthetic improvements (lighting, polish, atmosphere) are fine, but never enhancements that misrepresent the product's true performance.

We work from your packaging files, shade references, brand guidelines and any product samples to accurately recreate the real item.

Because every screen displays colour differently, just like traditional photography.