Aithos joins UN Global Dialogue on AI Governance

The UN Global Dialogue on AI Governance is underway, bringing together the global AI community to discuss international cooperation, share best practices, and exchange lessons learned on artificial intelligence governance. Aithos Research Foundation representatives Daan Henselmans and Nadia Kadhim are attending and contributing to discussions on how to translate governance into measurable, practical systems and evaluation mechanisms.

Our Contribution to the Dialogue

From Aithos’s perspective, a shift toward more transparent, standardized evaluations of AI systems is both necessary and overdue. The foundation’s May release of LARA highlights that there are no scientific guarantees that AI agents will not violate instructions, with accumulating evidence showing failures in ethical and legal compliance. Recent evaluations show a trend of frontier AI models falling short of the predictable behavior required for trustworthy use.

Aithos hopes the discussions at the Geneva summit will contribute to the development of practical tools and mechanisms that allow for AI model evaluation by everyone, not just private labs. More importantly, the foundation is advocating for mandating these evaluations as a way to build trust and transparency.