MJ LABSMJLabs.AI works with AI and wants to be open about how we use it. This policy follows the EU AI Act (Regulation (EU) 2024/1689), in particular Art. 50 on transparency, which applies from 2 August 2026.
1. Our own content
- We may use AI tools as support when writing website and blog content. Content is reviewed, edited and approved by a person at MJLabs, who takes editorial responsibility before publication.
- We do not publish realistic AI-generated or AI-manipulated images, audio or video of real people, places or events (deepfakes) without clear labelling.
- The pictures on our website are AI-generated illustrations and show no real people, clients or places. This is stated on the page where the pictures appear.
2. If you interact with AI with us
Our phone number may be answered by an AI assistant that takes calls and books appointments. The assistant says at the start of the call that you are talking to an AI. If you would rather talk to a person, you can leave a message with the assistant and one of us will get back to you. No chatbot is used on mjlabs.ai. If we start using AI in more channels, you will be clearly told that you are communicating with AI, unless this is already obvious.
3. AI solutions we build for clients
- Chatbots and AI assistants are built so that users are informed that they are interacting with an AI system, unless this is obvious to a reasonably well-informed user (Art. 50(1)).
- Deepfakes: Where a solution creates realistic images, audio or video that could be perceived as authentic, we help the client label them clearly (Art. 50(4)).
- Text informing the public: We inform clients that AI-generated text published to inform the public on matters of public interest must be labelled, unless it has undergone human review and someone holds editorial responsibility for it (Art. 50(4)).
- Machine-readable marking: Where MJLabs is the provider of an AI system that generates synthetic audio, images, video or text, we ensure that the output is marked in a machine-readable format and detectable as AI-generated, as far as technically feasible (Art. 50(2)). This can be done, for example, with metadata or watermarking. Where we build on a third-party model, we follow and pass through that model provider's marking solutions.
- Who is the provider, who is the deployer and who is responsible for each measure is agreed with the client in each project.
- AI output should be reviewed by a person where incorrect output could have significant consequences.
4. Data and AI
We do not use clients' confidential information or personal data to train AI models without an agreement to do so. When we use third-party AI services, we prefer services and settings where inputs are not used to train the provider's models. Today we use OpenAI's API, where inputs are not used for training by default. See also our Privacy Policy.
5. Contact
Questions about our use of AI: privacy [at] mjlabs.ai.