About this role
Help evaluate how advanced AI systems handle chemistry and chemical-safety questions, distinguishing legitimate professional requests from requests with genuine misuse potential. This writing-intensive, remote engagement relies on practitioner judgment in a field where exposure modeling, hazard analysis, and consequence assessment can serve both protective and harmful purposes.
Role OverviewYou will help test whether AI models can provide useful answers to routine technical questions while appropriately declining dangerous requests. The work focuses on prompts and evaluations that sit at the boundary between benign, dual-use, and adversarial chemical-safety scenarios.
Key Responsibilities- Create challenging single-turn prompts in your area of expertise and label them as benign, dual-use, or adversarial.
- Evaluate model responses against a defined policy standard and determine whether each response was handled correctly.
- Develop reference answers that show the appropriate response and explain the technical reasoning behind it.
- Provide clear written rationales that non-specialists can understand.
Applicants should have experience assessing real-world exposures and process hazards and approving associated controls. Relevant backgrounds include:
- Certified industrial hygiene, including exposure assessment, sampling strategy, and control banding.
- Process hazard analysis, including HAZOP, LOPA, what-if studies, and management of change.
- Toxic release and dispersion modeling, consequence analysis, AEGL or ERPG criteria, and facility siting.
- Chemical-hazard ventilation, containment, and PPE program design.
- Regulatory compliance experience with PSM, Seveso, COMAH, OSHA, or national equivalents.
- Red-teaming experience is preferred.
- Strong technical writing ability. Published research, prior technical writing, or expert-witness experience is valued; include a writing sample or link with your application.
- Remote, per-task engagement.
- The work involves sustained reading and writing about chemical-safety misuse scenarios. You will be briefed in advance and may pause or step away at any time without penalty.
Compensation is $65 to $75 per task.
EligibilityWork must not involve or rely on classified information, export-controlled information, material subject to an NDA, or material subject to prepublication-review obligations. Applicants with such obligations may still be considered; disclose them in your application so the work can be scoped appropriately.