AI & Equality Pub-Talk | Human Rights Benchmark for LLMs: Research Outcomes | Savannah Thais

We are advancing the Human Rights Benchmark for Large Language Models (LLMs)—a research initiative that examines how these systems align with core human rights principles. In this November Pub-Talk, Savannah Thais will present the outcomes of this work, sharing insights from the benchmarking process and highlighting what they reveal about the human rights implications of LLMs. This project reflects our commitment to building AI that respects dignity, advances equality, and serves all communities.

Why now? As LLMs are increasingly embedded in decision-making, chatbots, and public services, it is vital to move beyond accuracy toward accountability. Our research explores whether these models treat identities fairly, respond consistently to rights-based questions, and avoid harmful omissions or bias.