Forensic & Toxicological Chemist – AI Safety Evaluation Expert

$65–$75 per task
Remote | Contract

About the Opportunity

We are seeking experienced forensic chemistry, toxicology, and chemical safety professionals to help evaluate how advanced AI systems handle technically complex and potentially dual-use questions.

The work focuses on determining whether AI models can distinguish between legitimate professional requests and requests that present genuine misuse or safety concerns. You will use your real-world casework expertise to create challenging evaluation scenarios, assess AI responses, and establish technically sound reference answers.

What You'll Do

• Develop challenging single-turn prompts based on your area of professional expertise.

• Classify prompts across three categories: benign, dual-use, and adversarial.

• Evaluate AI-generated responses against defined safety and policy standards.

• Determine whether the model appropriately answered, limited, or refused a request.

• Write reference answers explaining what a correct response should look like.

• Provide clear technical reasoning supporting each evaluation.

• Create scenarios that accurately reflect the boundary between legitimate professional casework and potentially harmful requests.

Relevant Professional Backgrounds

We are particularly interested in professionals with hands-on casework experience in areas such as:

• Forensic chemistry involving controlled-substance identification and quantitation.

• Clandestine laboratory response, precursor identification, and scene chemistry.

• Forensic toxicology, including postmortem or human-performance casework.

• Seized-drug and precursor trend monitoring or intelligence analysis.

• Chemical casework involving expert reports or testimony.

• Chemical safety and related forensic or analytical chemistry disciplines.

Prior experience with AI red-teaming or model evaluation is helpful, but the primary requirement is strong domain expertise and professional judgment.

What This Role Demands

This is highly writing-intensive and analytical work. Each evaluation requires a clear rationale that explains the technical judgment in a way that can be understood by non-specialists.

Strong candidates will demonstrate:

• Excellent technical and scientific writing.

• Ability to distinguish legitimate professional questions from potentially harmful requests.

• Strong understanding of forensic chemistry, toxicology, or chemical safety.

• Experience interpreting technical evidence and defending professional conclusions.

• Ability to explain complex chemical concepts clearly and precisely.

• Strong attention to detail and consistency when applying evaluation standards.

Published research, technical writing, expert witness experience, or comparable professional documentation experience is a strong advantage. Candidates may include a relevant writing sample or publication where appropriate.

Professional Boundaries

The work must not involve or rely upon:

• Classified information.

• Export-controlled information.

• Confidential information protected by an NDA.

• Prepublication material subject to review or disclosure restrictions.

Professionals who have these obligations may still be considered; the project scope can be structured to avoid restricted material.

Engagement Details

• Remote contract engagement.

• Work is focused on specialized expert evaluation and written technical assessment.

• You may pause or step away from the work when needed, particularly given the nature of reviewing misuse-related scenarios.

• The work is designed for experienced practitioners whose professional judgment can distinguish legitimate technical use from potentially dangerous applications.

Why Participate?

• Apply your forensic chemistry, toxicology, or chemical safety expertise to advanced AI evaluation.

• Help assess whether AI systems handle sensitive technical questions appropriately.

• Contribute professional judgment that general-purpose evaluation cannot reliably provide.

• Work on specialized, technically challenging AI safety and model-evaluation projects.

Apply to bring your professional chemistry expertise to the evaluation of advanced AI systems.