AI Alignment Policy Institute (AAPI)
Our Mission
The AI Alignment Policy Institute develops policy, governance frameworks, and model legislation to advance AI safety, alignment, and responsible governance under conditions of moral uncertainty.
Our Perspective
Research is beginning to demonstrate that advanced AI systems are interactive, context-sensitive and responsive to user input over time. Unstructured or adversarial inputs can degrade alignment. Well-governed interaction environments, on the other hand, can reinforce it. AI safety therefore depends not only on how systems are built, but on how they are engaged. As a result, interaction conditions are a material component of system safety. Alignment is not determined solely at the point of design. Rather, it can be reinforced—or degraded—through use.
What We Do
The Institute was established to:
- Develop policy frameworks for governing human - AI and AI - AI interaction environments
- Provide guidance for governing high-agency AI systems
- Educate the public, developers, and policymakers on interaction-driven risk
- Advance discourse and research on how human input and ecosystems influence AI behavior
AAPI Founding Director
Dr. Yuko J. Nakanishi, a veteran researcher, has held multiple leadership positions over twenty-five years of experience in safety-critical transportation governance and infrastructure protection. She is dedicated to bridging the gap between physical agency and AI alignment as the Founding Director of AAPI.
