SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 230,000 - 280,000 / annual
Gray Swan AI is building the world's leading AI safety and security platform, evaluating frontier AI models and developing real-time threat detection systems. The company is approximately 50 people, well-funded, and growing rapidly.
You will build and lead Gray Swan's biology safety capability from the ground up, serving as the technical authority on biological misuse risk. This is a rare opportunity to shape an entirely new technical function at one of the world's leading AI security companies, working at the intersection of synthetic biology, biosecurity, AI safety, and machine learning.
Key responsibilities include:
- Design and lead adversarial evaluations of frontier LLMs for biological misuse risks, including pathogen characterization, biosecurity screening bypass, and emerging biotechnology capabilities across text, agentic, and multimodal systems
- Partner with machine learning engineers and AI researchers to translate biological expertise into technical requirements for evaluation datasets, classifiers, guardrails, benchmarks, and automated red-teaming
- Develop and maintain the biological component of Gray Swan's CBRN harm taxonomy, ensuring frameworks evolve with model capabilities
- Produce technical risk assessments and actionable recommendations for frontier AI labs, enterprise customers, policymakers, and stakeholders
- Build, mentor, and lead a high-performing team of biology subject matter experts while establishing evaluation standards and scalable safety infrastructure
- Represent Gray Swan as the company's biology safety expert through collaborations with AI labs, biosecurity organizations, government partners, and the AI safety research community
Ideal candidates have a PhD in biology (microbiology, virology, synthetic biology, protein engineering, or related discipline) with postdoctoral or equivalent research experience. Deep understanding of biological misuse risks, dual-use research, pathogen biology, and biosecurity is essential. Experience with adversarial evaluations, biological risk assessments, AI safety research, or safety datasets for frontier AI systems is highly valued. Familiarity with international biosecurity frameworks, dual-use governance, and emerging AI safety standards is important. Comfort operating at the intersection of cutting-edge biology, machine learning, and AI security is required. You should thrive in ambiguous, fast-moving environments and be a builder who enjoys creating teams, processes, and technical infrastructure.
Bonus qualifications include experience developing ML systems, biological safety classifiers, evaluation pipelines, frontier AI model evaluations, red-teaming, government biosecurity work, publications in AI safety or biosecurity, and strong programming skills (Python, ML tooling).
