Animals-Aligned Agentic AI Benchmarking

AI/ML Research -
AI Safety Agentic AI Animal Welfare Benchmarking Futurekind

As a Futurekind Research Fellow with Electric Sheep, I worked on benchmarking whether implicit values in agentic LLMs are inclusive of animals.

The project treats animal inclusion as an evaluation problem for agentic AI systems: when agents reason about plans, tradeoffs, stakeholders, and consequences, do animal interests appear in their implicit value model or get ignored by default?