Enabling Safer AI
Our Research

Enabling Safer AI Through Smarter AI.

We tackle some of the most pressing issues of current AI, holding back its potential for societal and commercial value in "cannot-fail" industries such as medicine, law, and finance.

Sculptural profile figure with geometric triangle overlay on archival paper

{{ i.title }}

{{ i.description }}
Recent Work

We focus on the foundational issues that are the root causes of today's safety-related issues — AI systems must understand what they read, what they see, and the implications of their actions, and must be designed with humans in mind.

{{ group.group }}

{{ it.title }}

{{ it.body }}

{{ it.cite }}
{{ it.cite }}

{{ it.body2 }}

{{ it.cite2 }}
{{ it.cite2 }}
Philosophy

Deep, long-term, multi-disciplinary programs.

{{ ph.title }}

{{ ph.description }}

Is this up your alley?

Collaborating Internships Hiring