Researcher (AI Safety)

TL;DR

Researcher (AI Safety): Designing and implementing mitigation components to reduce loss of control risks in advanced AI systems with an accent on safeguards, monitoring, detection, and enforcement. Focus on building robust protections against misaligned, deceptive, or uncontrollable model behaviors in frontier AI models.

Location: Onsite in San Francisco, USA

Salary: $295,000–$445,000

Company

hirify.global is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity by safely developing and deploying advanced AI systems.

What you will do

• Design and implement mitigation components for loss of control risk including prevention, monitoring, detection, containment, and enforcement.
• Integrate safeguards across product and research teams ensuring protections are consistent, low-latency, and resilient.
• Evaluate technical trade-offs and propose pragmatic, testable solutions within the loss of control domain.
• Collaborate with risk modeling, evaluations, and policy partners to align mitigation design with high-severity threat scenarios.
• Execute rigorous testing and red-teaming workflows to stress-test mitigation stacks against subversive model behaviors.

Requirements

• Location: Must be onsite in San Francisco, USA
• Passion for AI safety and experience with deep learning and transformer models.
• Proficiency with PyTorch or TensorFlow frameworks.
• Strong foundation in data structures, algorithms, and software engineering principles.
• Experience designing and evaluating technical safeguards and control mechanisms for advanced AI behavior.
• English: C1+ proficiency required

Nice to have

• Background knowledge in alignment, control, interpretability, robustness, adversarial ML, or related fields.

Culture & Benefits

• Equal opportunity employer valuing diverse perspectives and experiences.
• Commitment to reasonable accommodations for applicants with disabilities.
• Work on cutting-edge AI safety challenges with societal impact.