AI TRANSPARENCY COMPLIANCE GROUP

Ensure compliance with Article 50(2) of the EU AI Act

Mark your content and use AI detection to facilitate authentication.

Context

The EU AI Act

Starting in August 2026, providers of generative AI systems available in the European Union will be required to label and make their AI-generated content detectable.

Article 50(2) of the European AI Regulation (also known as the EU AI Act) requires multi-layer marking of generated content as well as the provision of a public detector on their website.

Label4.ai and KeeeX are pooling their expertise to offer you a sovereign, turnkey solution that guarantees immediate compliance with European regulations. Failure to comply may result in a fine of €15 million or 3% of global revenue.

The European AI Regulation

The Consortium

A combination of cutting-edge sovereign solutions

Label4.ai is a cutting-edge European solution developed through public research that enables the application of an invisible and robust digital watermark to all types of AI-generated content (images, videos, audio, documents, and text). The company also offers advanced, sovereign forensic detection technology.

https://label4.ai

KeeeX is a pioneer in secure metadata and has been developing software for digital trust and security since 2014. Its universal, scalable, and resource-efficient process makes it easy to sign, mark, track, and verify any digital file without requiring any infrastructure. Users can opt to include an AI summary, a prompt, or an opt-out option in the metadata.

https://keeex.me

SOLUTION

A compliance solution consisting of two mandatory components:

Invisible Watermarking – Label4.ai

Integration of an invisible watermark directly into the content that is resistant to alterations (cropping, compression, rotation). It ensures the watermark remains intact and allows for the long-term identification of the file’s origin or indicates the presence of secure metadata within the file.

Secured Metadata – KeeeX

Adding secure metadata to the file to verify its integrity, origin, and date. Lightweight and universal, this metadata ensures the robustness of the marking and allows for the addition of free-form references (tags, versions, links, license, copyright, classification, AI prompt, opt-out, etc.).

And another building block:

Forensic Detection – Label4.ai

Enables the detection of files previously marked by Label4.ai and files without active tags. This detection provides a probability score for AI-generated content within the files.

Simple. Secure. Scalable.

Our Commitments

Available via API

Quick and easy API integration for seamless integration with your tools

Local deployment is possible

On-premises deployment on demand for local, ultra-fast data processing

Integration into generation pipelines

A solution specifically designed for labeling content from AI system providers

Data Privacy

Your data is never stored and can be processed locally

Compliance Guaranteed

We ensure your compliance with the Code of Practice, which we helped draft

Turnkey Sovereign Solution

Developed by two French and European companies, with no reliance on offshore entities

Benefits

Protect your brand and intellectual property

Marking, signing, and timestamping all your intangible assets so that anyone can verify that they originate from you and prevent them from being misappropriated or tampered with.

Minimize the risks associated with deepfakes and document fraud

Protecting your business and your employees is crucial. AITCG helps you screen incoming content and flag outgoing content to facilitate its review.

Strengthen your cybersecurity and build trust within your ecosystem

Ensuring transparency and trust in systems that rely on third-party components and AI-generated content has become essential with the rise of synthetic content.

Enhance your offering and stand out from your competitors

By integrating our proprietary solution, you can offer your customers a unique service and set yourself apart from your competitors.

ECOSYSTEM

Integrators, Suppliers, Deployers, Users

Join us



"Providers of AI systems, including general-purpose AI systems, generating synthetic audio, image, video or text content, shall ensure that the outputs of the AI system are marked in a machine-readable format and detectable as artificially generated or manipulated. Providers shall ensure their technical solutions are effective, interoperable, robust and reliable as far as this is technically feasible, taking into account the specificities and limitations of various types of content, the costs of implementation and the generally acknowledged state of the art, as may be reflected in relevant technical standards."

– Article 50(2) of EU AI ACT –

"Such techniques and methods should be sufficiently reliable, interoperable, effective and robust as far as this is technically feasible, taking into account available techniques or a combination of such techniques, such as watermarks, metadata identifications, cryptographic methods for proving provenance and authenticity of content, logging methods, fingerprints or other techniques, as may be appropriate."

– Recital 133 –

Contact

Ask for a demo

Would you like to try out our solutions or need more information? Feel free to contact us!