Key takeaways from the Stibbe Tech Seminar

Saskia Lensink gave the keynote speech and introduced GPT-NL, the Dutch large language model built on reliable data and European values. She shared insights on fair copyright and the TNO initiative. Jan-Jaap Koningsveld discussed the rapid legal developments in Dutch and European AI law. This was followed by a panel discussion on practical AI compliance with Marko Jovović, Auke-Frank Tadema and Minke Reijneveld, moderated by Jan-Jaap.

Key take-aways:

• TNO's Saskia Lensink delivered a powerful wake-up call: while commercial AI models often train on murky, undisclosed datasets, initiatives like GPT-NL are charting a different path. This Dutch Large Language Model demonstrates that organizations can develop powerful AI systems using transparent, lawfully obtained datasets while maintaining compliance with copyright and privacy requirements from the outset.
• GPT-NL is digital sovereignty in action, demonstrating that European AI development can successfully balance technological innovation with regulatory compliance and ethical standards.
• Developments around the E.U. AI Act emerge daily. As further complexities become apparent, the need for comprehensive legal guidance becomes essential.
• Notwithstanding ongoing regulatory developments, organizations can implement foundational compliance measures immediately:
  • Conduct comprehensive AI system inventory and risk assessment
  • Implement AI governance training programs across all organizational levels
  • Establish comprehensive documentation protocols for data sources, training methodologies, and model development processes
  • Integrate ethical review processes and governance frameworks throughout the AI development lifecycle
  • We look back on a very interesting, informative and successful meeting. If you have any questions about the seminar or tech-related topics, please contact Pieter Schütte.