Biologist for AI Safety Evaluation

Apply PhD-level biology expertise to evaluate and improve how advanced AI models handle specialized life-science content. In this role you will craft expert prompts, assess model outputs for scientific correctness and safety, and follow structured classification guidelines to reduce risk. No prior AI or machine learning experience is required, training on the workflow will be provided.

• Write expert-level prompts across specialized life-science topics.
• Evaluate and annotate model responses for scientific accuracy, helpfulness, and appropriate handling of sensitive content.
• Apply structured guidelines to classify prompts and conversations.
• Use your scientific judgment to identify potential risks and suggest improvements to model behavior in life-science contexts.

• PhD in biology or a closely related life-science field, ongoing or completed, for example molecular biology, microbiology, virology, genetics, biochemistry, bioinformatics, synthetic biology, or immunology.
• Deep familiarity with modern laboratory and computational techniques relevant to your subfield.
• Strong scientific reasoning and clear scientific writing in English.
• Sound judgment regarding biosecurity and the responsible handling of dual-use information.

• Research or coursework touching on biosafety, biosecurity, pathogen biology, or gain-of-function considerations.
• Experience reviewing, grading, or red-teaming technical content.

• Start date: Immediate.
• Commitment: Part-time, 15 to 25 hours per week, with flexibility to work up to 40 hours per week.
• Employment type: hourly.

$65 to $70 per hour.

Candidates must be located in the United States.