Safety by Design for AI Systems
Applying behavioral science to build AI systems that are safe, trustworthy, and human-centered from the start.
Partners
Status
Status
Ongoing
Ongoing
Focus
Focus
AI Safety
AI Safety
Overview
Safety should not be an afterthought. As AI systems become increasingly integrated into everyday life, safety must be considered from the earliest stages of design and development.
This project develops Safety by Design principles that integrate behavioral science into the AI development lifecycle. Rather than responding to harms after deployment, the research focuses on preventing risks before they emerge.
Research Approach
The project combines behavioral science, Trust & Safety, and human-centered AI to identify how AI systems can unintentionally influence human trust, decision-making, and behavior.
The project explores practical design strategies that reduce manipulation, harmful engagement, addictive interaction patterns, and other behavioral risks while promoting transparency, appropriate trust, user autonomy, and digital well-being.
Expected Inpact
The project enables technology companies to integrate behavioral safety into the design of AI-powered products and digital platforms. By addressing behavioral risks early in development, organizations can build experiences that foster healthy engagement, strengthen user trust, and reduce the likelihood of manipulation, harmful interactions, and unintended negative outcomes.
RESEARCH PROJECTS
Research at the Intersection of Human Behavior and AI.
Safety by Design for AI Systems
Applying behavioral science to build AI systems that are safe, trustworthy, and human-centered from the start.
Partners
Status
Status
Ongoing
Ongoing
Focus
Focus
AI Safety
AI Safety
Overview
Safety should not be an afterthought. As AI systems become increasingly integrated into everyday life, safety must be considered from the earliest stages of design and development.
This project develops Safety by Design principles that integrate behavioral science into the AI development lifecycle. Rather than responding to harms after deployment, the research focuses on preventing risks before they emerge.
Research Approach
The project combines behavioral science, Trust & Safety, and human-centered AI to identify how AI systems can unintentionally influence human trust, decision-making, and behavior.
The project explores practical design strategies that reduce manipulation, harmful engagement, addictive interaction patterns, and other behavioral risks while promoting transparency, appropriate trust, user autonomy, and digital well-being.
Expected Inpact
The project enables technology companies to integrate behavioral safety into the design of AI-powered products and digital platforms. By addressing behavioral risks early in development, organizations can build experiences that foster healthy engagement, strengthen user trust, and reduce the likelihood of manipulation, harmful interactions, and unintended negative outcomes.
RESEARCH PROJECTS