AI Transparency

At Koolbrand we use artificial intelligence. And we say so.

AI is one more tool in our creative and growth operation, like Figma, Photoshop or a decent coffee. It helps us explore faster, validate better and create with less friction. But the brand decision, the creative direction and the editorial responsibility are always human.

This page explains how we do it, so that clients, team and authorities know exactly what to expect.

What we count as AI-generated content

We label as “AI-generated” any content that meets both conditions:

1. It comes out of a generative model (text, image, audio or video).
2. It is published without substantial human alteration that changes its meaning.

If someone on the team reviews it, edits it, signs it and takes editorial responsibility, the content becomes human-assisted work and needs no label, as set out in the European AI Act (Art. 50.4).

How we label

We apply a two-layer system across every channel:

Visible layer

• Images: an “AI-generated” badge in the footer or caption.
• Fully AI video: a notice visible throughout playback.
• Video with AI segments: a notice at the start.
• Synthetic voices: disclosure at the beginning of the audio.
• Chatbots and assistants: a notice reading “You are talking to an AI assistant” before the first message.

Invisible layer (machine-readable)

Standard C2PA metadata embedded in every file, so platforms and verification tools detect it automatically.

The AI tools we use

These are the platforms we work with today. We review them every quarter.

All of these are third-party tools subject to their own terms. Our use is governed by our Internal AI Policy (available on request).

Our human editorial process

Every piece we publish goes through at least one of these filters before it ships:

1. Creative review: art direction or copy checks it against the client account.
2. Strategic review: account or brand strategist confirms it meets the brief.
3. Client approval: the client signs off the delivery before publication.

Once a piece has passed these filters and a human takes editorial responsibility, it falls under the exception in Art. 50.4 of the European AI Act and requires no mandatory label, though we may label it voluntarily where that adds clarity.

For Koolbrand clients

If we work with you, this is what we agree by default on any project involving AI:

• We will tell you which AI tools we will use on your project.
• We will label the deliverables that legally require it.
• We will agree who, Koolbrand or you, is responsible for keeping the labels in place at publication.
• We will document the use of AI at the close of each project.

This is set out in a specific clause in the contract. If you have questions, write to us at hola@koolbrand.com.

Applicable legal framework

• Regulation (EU) 2024/1689 — the European Artificial Intelligence Act. The transparency obligations in Art. 50 apply from 2 August 2026, with an adaptation period until 2 December 2026.
• AESIA — the Spanish Agency for the Supervision of Artificial Intelligence, based in A Coruña. The competent national authority.
• Code of Practice on Transparency of AI-Generated Content (European Commission, 10 June 2026).
• Official EU icons — digital-strategy.ec.europa.eu.

Contact and review

This policy is reviewed every 6 months or whenever new legislation comes into force.

AI editorial lead: Koolbrand management team
Contact: hola@koolbrand.com
Last updated: 30 June 2026
Next review: December 2026

If you think any of our content should be labeled and is not, or you have any question at all, write to us. We always reply.