Research Manager, Biological Safety
$405k - $485k • Remote • San Francisco, CA
Posted 1mo ago
Job Location
San Francisco, CA
Tech Stack
Remote Work Policy
Fully remote
Categories
Machine Learning Engineer
About the job
Anthropic's Safeguards organization is responsible for developing the policies, evaluations, and enforcement systems that prevent AI models from causing catastrophic harm. We are seeking a manager to lead our biological safety research engineering team. This team focuses on creating the evaluations, datasets, and classifiers that govern how our models handle biological knowledge. You will guide a team of research scientists and engineers in designing and executing capability evaluations on frontier models, curating safety classifier training data, and training/iterating on these classifiers. Your role involves setting technical direction, making investment decisions, and owning team outcomes, while maintaining technical depth to review designs and analyze failure modes. The team's core challenge is balancing robust safeguards against sophisticated actors with enabling legitimate researchers in the life sciences.
Responsibilities
- Manage, coach, and grow a team of research scientists and engineers focused on biological safety evaluations and classifiers, including hiring, onboarding, performance, and career development.
- Set the technical direction and roadmap for biological safety research, and make decisions on team priorities and safeguard deployment.
- Own the quality of capability evaluations assessing model abilities in the biological domain and translate results into actionable deployment recommendations.
- Guide the development of training and evaluation datasets for safety classifiers, collaborating with threat modeling experts.
- Oversee the training and iteration of safety classifiers with ML engineers, optimizing for adversarial robustness and low false-positive rates.
- Ensure the team invests in tooling and pipelines for efficient and repeatable evaluation and classifier development.
- Establish metrics for measuring classifier and evaluation performance against production traffic, identifying gaps, and prioritizing improvements.
- Direct red-teaming and stress-testing of safeguards against evolving threats, models, and product surfaces.
- Partner with Research, Product, Policy, and government affairs teams to integrate biological safety into the model development lifecycle and act as an escalation point for biological content.
- Represent the team's work in external communications, including model cards, blog posts, and policy documents.
- Track advancements in biology, machine learning, and biosecurity for potential risks and mitigation opportunities.
Requirements
- Experience managing a technical team, including hiring, coaching, and performance management.
- Proven ability to set technical direction for a team and make prioritization decisions under uncertainty.
- Proficiency in Python with a background in scientific programming and data analysis.
- Solid understanding of ML fundamentals to critically review evaluation design and classifier development.
- Knowledge of modern biology across measurement and engineering, including high-throughput assays, functional characterization, gene synthesis, genome editing, strain construction, and protein engineering.
- Experience designing quantitative experiments or evaluations and drawing defensible conclusions from noisy results.
- Clear analytical and writing skills, with the ability to explain technical concepts to non-technical stakeholders.
- Familiarity with dual-use research concerns and biosecurity frameworks (e.g., select agent regulations, Biological Weapons Convention, Australia Group guidelines).
- Comfort with ambiguity and shifting priorities in a rapidly evolving AI landscape.
- Motivation to prevent misuse while enabling beneficial applications in the field.