Biosecurity & CBRNE Subject Matter Expert - AI Safety-; Remote

A Global Leader in AI Development

Full time

Location

United States (Remote)

Work environment

Remote

Expected pay amount

65.00 USD Per Hour

Schedule

Flexible:
Monday-Friday, approximately 40 hours per week (whatever timezone the candidate is in)

Contract

Hire Art is helping our client find a Biosecurity & CBRNE Subject Matter Expert - AI Safety to support the safe development and evaluation of next-generation AI systems, with a focus on chemical, biological, radiological, nuclear, and explosive (CBRNE) risks.

In this role, you'll apply your scientific, technical, or operational expertise to red-team advanced AI models, identify potential safety vulnerabilities, and evaluate their ability to generate harmful or high-risk information. You'll develop adversarial prompts to test AI safeguards and assess model outputs for accuracy, specificity, actionability, and whether they provide meaningful information beyond what is already available through published literature and other public sources.

The ideal candidate brings hands-on scientific or operational experience, strong analytical judgment, and an interest in addressing emerging challenges at the intersection of AI, biosecurity, and public safety.

• Develop adversarial prompts to test AI model safeguards and identify potential safety vulnerabilities.
• Evaluate AI-generated outputs for scientific accuracy, specificity, actionability, and the degree of meaningful uplift they provide beyond publicly available information.
• Apply specialized biological or CBRNE expertise to assess complex technical information and identify concerning AI model capabilities related to chemical, biological, radiological, nuclear, or explosive hazards.
• Determine whether AI-generated information lowers practical barriers related to knowledge, materials, equipment, or skills, or simply restates publicly available information.
• Contribute to AI safety benchmarks, risk thresholds, severity rubrics, and evaluation frameworks while calibrating assessments with other subject matter experts.
• Document findings and provide clear rationales for severity assessments to support actionable improvements by AI developers.
• Collaborate with cross-functional teams across AI safety, research, security, and policy.

• Deep technical expertise in at least one CBRNE domain (biological, chemical, radiological/nuclear, or explosives), demonstrated through a graduate degree with hands‑on laboratory or field experience in a relevant discipline, or equivalent operational experience (e.g., CBRN specialist, EOD technician, HAZMAT, biodefense, or national laboratory work)
• Strong scientific and analytical judgment, including the ability to distinguish widely available information from information that could materially lower barriers to causing harm
• Excellent attention to detail and ability to identify errors, inconsistencies, and potential safety concerns
• Strong written communication skills and ability to clearly document technical findings
• Familiarity with prompting or interacting with large language models (LLMs); prior AI red-team experience is not required
• Interest in AI safety and the potential risks associated with advanced AI capabilities
• Ability to handle sensitive or potentially hazardous information responsibly while following confidentiality and information-handling protocols
• Must be a U.S. citizen

• A PhD, MD, or DVM in a relevant scientific, medical, veterinary, or technical field
• Military experience involving CBRN, CBRNE, or EOD operations, such as a CBRN specialist, EOD technician, or related role
• Experience in HAZMAT, emergency response, biodefense,…