project Coming soon
AI Agent Assessment
Assess Agentic AI for their ability to perform tasks appropriately for the public.
Our Research
DSRI is researching methodologies to advance public safety assessment of rapidly evolving AI/digital systems.
project Coming soon
Incident Assessment
Assess public harm incidents associated with specific types of AI/digital systems.
project Feb 2025 — Ongoing
Content Assessment
Assess synthetic and manipulated media detection tools for their ability to properly explain their function, outputs, etc., so that users can properly use these imperfect tools.
project Jun 2023 — Dec 2025
Chatbot Assessment
DSRI is assessing LLMs for their ability to provide expert advice in medicine, software development, and therapy.
partnership Jan 2022 — Dec 2025
Center for Advancing Safety of Machine Intelligence (CASMI)
A partnership between ULRI and Northwestern University to establish best practices for the evaluation, design, and development of machine intelligence that is safe, equitable, and beneficial.