The transition from research to application is often where AI projects prove their true value. For GPT-NL, that moment has arrived. A select group of organizations—(including government bodies and knowledge institutions) are now testing early versions of the model in their own environments.

These so-called “feasibility studies” are designed to evaluate how GPT-NL performs in real-world contexts: from technical infrastructure and performance to alignment with specific organizational use cases. Early applications include supporting public service chatbots, assisting with document generation, and analyzing large datasets in sensitive domains.

What sets GPT-NL apart is its focus on sovereignty and trust. The model is developed using legally sourced, verifiable datasets and is designed to run within secure, local infrastructures. This allows organizations to maintain control over their data (an increasingly critical requirement in sectors where privacy and compliance are non-negotiable).

The initiative is backed by public investment and aims to reduce dependency on non-European AI providers. By offering an auditable and transparent alternative, GPT-NL contributes to a broader movement toward responsible and accountable AI ecosystems in Europe.

Importantly, this pilot phase is not just about validating technology, it is about learning. Feedback from the first five customers will directly inform further development, helping shape the model into a robust tool for organizations across sectors.

As the project moves toward broader availability later in 2026, GPT-NL represents more than a technical achievement. It signals a shift toward AI that is built with public values at its core, balancing innovation with ethics, and capability with control.