In January we discussed the challenges and opportunities regarding AI usage among the youth, in March we delved into the future of AI regulation, and in May we debated on whether LLMs can reason, and why that matters.
AI and youth
28 January 2026 – with Nic Orchard (TU Delft), Dr. Rebecca Wald (Netherlands Youth Institute), and Dr. Elske van den Hoogen (University of Amsterdam), moderated by Human(e) AI postdoctoral researcher Zilin Lin
Nic Orchard focused on mental health, using intersectional feminist design and participatory methods to create more equitable AI-based digital mental health interventions. Rebecca Wald shared her insights into growing up and parenting in a digital world, particularly how families adopt and use modern (AI) technologies such as virtual assistants. Our own Human(e) AI postdoctoral researcher Elske van den Hoogen presented the project “Disinforming
YOU(th)”, which investigates how Dutch adolescents perceive and create AI-generated disinformation.
Together, this session covered smart devices, chatbots, metal health, and the creation of safer online environments for the younger generation.
The Future of AI Regulation in the EU
3 March 2026 – with Prof. Dr. Natali Helberger (University of Amsterdam) and Prof. Dr. Hannah Ruschemeier (University of Osnabrück), moderated by Human(e) AI postdoctoral researcher Ljubiša Metikoš
Hannah Ruschemeier and Natali Helberger discussed the various legislative changes proposed in the Digital Omnibus Package. Potentially, the Omnibus could deeply impact the entirety of the EU Digital Acquis. In this session, we focused on the proposed changes to key definitions in the GDPR as well as various aspects of the AI Act. While these proposed changes aim to simplify EU regulation and improve Europe’s competitiveness, the question was raised whether this goal would truly be achieved with the Omnibus Package. The potential danger of weakening Union citizens’ rights was also highlighted by the discussants.
Can LLMs reason?
28 May 2026 – with Prof. dr. ir. Jan Broersen (Utrecht University) and Prof. dr. Bart Verheij (University of Groningen), moderated by Human(e) AI postdoctoral researcher Mina Young Pedersen
Bart Verheij and Jan Broersen shared their perspectives on what reasoning means in the context of LLMs. There are many ways to distinguish different types of reasoning, dividing it into normative (how one ought to reason) and descriptive (how we humans actually reason) reasoning; deductive (derive conclusions from premises), inductive (derive general statements from examples) and abductive (derive an explanation from observations) reasoning; or a static versus an agentic view of reasoning, where the latter involves a dynamic process of doing the reasoning. Bart and Jan argued that LLMs are worse at doing the more typically human types of reasoning such as abductive and agentic reasoning. They also discussed the future of AI technology, whether we can have LLMs that reason and whether this technology can be useful. A point was made that in the development of ever more advanced AI technology, we will find back to our humanity and cherish what is inherently human.