AI Safety Researcher

About the Role

Cohere is building an AI Safety Research team to ensure our enterprise models are safe, reliable, and aligned with human values at scale.

Research Areas

• Robustness and adversarial attacks on enterprise LLMs
• Hallucination detection and mitigation
• Bias and fairness in domain-specific fine-tuned models
• Red-teaming and evaluation methodologies

Why Cohere

We're at the intersection of safety research and real-world enterprise deployment, giving you unique insight into how safety challenges manifest in production.

• PhD in ML, AI Safety, or related field
• Strong publication record in safety, robustness, or alignment
• Experience with red-teaming or adversarial ML
• Excellent communication skills

• Fully remote globally
• Top-tier compensation + equity
• Research publication support
• Conference travel budget