Curated roles
Browse roles focused on reducing risks from advanced AI systems, including alignment research, safety engineering, evaluations and related operations.
357 active roles found.
Generalist role supporting Kairos’s AI safety university groups program, including the Pathfinder Fellowship, organizer support, events, and program infrastructure.
Research role focused on mechanistic interpretability and understanding model internals to improve alignment and safety of powerful AI systems.
Analyst role focused on enforcing safeguards against misuse of Anthropic's AI systems for conventional weapons and dangerous technology, including model behavior assessment, enforcement workflows, and evals.
Engineer role focused on cyber-relevant model evaluations, safeguard robustness testing, and misuse detection for frontier AI systems.
Research scientist role on Apollo Research’s AI control and monitoring team, designing control protocols, evaluation frameworks, and monitors to reduce risks from AI systems and coding agents.
Security and control researcher for coding agents, focused on AI threat modeling, red-teaming, and improving monitors and control protocols to reduce catastrophic risks from misaligned or compromised agents.
Dedicated AI red team engineer role focused on red-teaming AI monitors, finding attack surfaces, and improving safety monitoring for coding agents and frontier lab systems.
An intensive, fully-funded ten-week AI safety research fellowship in Cambridge focused on technical AI safety, governance, interpretability, formal verification, model evaluations, and risk management frameworks.
Security research role focused on AI auditing methodologies, AI system vulnerabilities, and tools/frameworks for evaluating AI safety and security.
Lead enforcement operations to detect and mitigate misuse of Anthropic AI systems for cyberattacks, malware, and offensive exploitation, including managing a team and shaping enforcement strategy.
Frontier AI red teaming role focused on attacking cyber safeguards, designing attack strategies, and feeding results into safety/alignment evaluations and deployment decisions.
Part-time technical advisor role for AI safety research on fine-tuning and evaluations, advising on benchmark design, generalization testing, and validity of welfare/capability evals for open-weight models.
Technical security researcher role focused on red-teaming AI agent systems, securing agent sandboxes and access controls, and building monitoring and controls to reduce catastrophic risks from coding agents.
Research engineer role on DeepMind’s AGI Safety and Alignment Team, focused on alignment methods, adversarially robust control systems, interpretability, and frontier-model safety work.
Research scientist role leading Frontier Safety Framework governance research and risk modeling for frontier AI models, including safety evaluations, mitigation assessment, and external safety reporting.
A 3-month full-time research fellowship in applied mathematics for AI safety, centered on technical AI alignment research and producing a research proposal.
Research scientist role focused on monitoring deployed GenAI models for safety, alignment, misuse, and coordinated harms using automated evaluations, classifiers, and large-scale production data.
Research engineer role on DeepMind's AGI Safety and Alignment Team focused on alignment methods, adversarially robust AGI control systems, and interpretability for frontier models.
Senior Model Policy Manager role focused on creating and operationalizing policies, taxonomies, and evaluation criteria to keep frontier AI systems safe, especially for biological and chemical risk.
Founding grantmaker for the AI Model Safety program, funding independent evaluations, standards, and foundational safety research to improve frontier model safety.