AI Behavior Researcher - Child Safety and Mental Health

Salary range: $250,000 - $500,000/year + benefits


Description: Transluce is a fast-moving nonprofit research lab building the
public tech stack for AI evaluation and oversight. We are pioneering research
into how AI chatbots and companions impact mental health and child safety, and
we’re improving outcomes for millions of sensitive AI interactions with
vulnerable users.


About the role: We are looking for interdisciplinary researchers who can bridge
domain expertise in mental health, child safety, or adjacent social science with
enough technical fluency to actively develop and refine evaluation methods. You
don't need to be a pure ML engineer, but you should be comfortable working
inside existing evaluation pipelines: adapting user simulators, refining judge
prompts and rubrics, and collaborating with external domain experts to validate
what you're measuring.


As an early member of a highly collaborative team, you will learn and grow
quickly, working directly with leading AI researchers, frontier AI labs, and
prominent child safety and mental health experts.


Core responsibility: Build and extend Transluce’s AI evaluation methods and
oversight tools for mental health, child safety, and related topics. This
includes:
* Identify and measure important emerging AI behaviors that may impact
vulnerable users.
* Identify gaps in current evaluations and design methods to address them.
* Adapt existing methods to cover new domains, monitor evolving safety issues
and trends, and inform key decisions in industry and government.
* Carry out human subjects research to inform and validate evaluation design.
* Build and manage relationships with clinicians, social science researchers,
affected users, civil society, and relevant safety researchers at frontier
labs. 



Qualities of a strong candidate:
* Quantitative research background in AI evaluation, HCI, psychology, social
data science, public health, or a related field.
* Enough ML and programming fluency to navigate and modify existing evaluation
codebases, even if you wouldn't build infrastructure from scratch.
* Reliable and trustworthy results: meticulous, good experimental design,
epistemic self-awareness and transparency.
* Ability to balance between the needs of AI researchers and domain experts, as
well as between researchers and senior decision makers.
* We're particularly interested in candidates with hands-on experience working
directly with domain experts to integrate their expertise into system design.



We are located in San Francisco and enthusiastic to work together in-person. We
are open to sponsoring international visas.