Curated roles
Browse roles focused on reducing risks from advanced AI systems, including alignment research, safety engineering, evaluations and related operations.
393 active roles found.
Remote research management role mentoring AI safety fellowship mentors, supporting research leads, team management, and research quality across SPAR.
Research Engineer on the Eval Platform team building infrastructure for reproducible, scalable model evaluations and dashboards used by researchers to assess model quality.
Research engineer role focused on building and operating SimpleAudit for AI safety audits, red teaming, and safety evaluations of language models and AI systems.
Senior engineering leadership role building the engineering systems, tools, and team for FAR.AI’s frontier AI red-teaming program, including safety evaluations, attacker simulation, and harmfulness evaluation for frontier models.
Project Officer role focused on building a secure cyber-range to evaluate AI agents and frontier models in cybersecurity, including red-team/blue-team scenarios and vulnerability testing.
Senior TPM role on DeepMind's AGI Safety and Alignment team, driving technical execution for Gemini safety, evaluation infrastructure, monitoring, and risk mitigation across the model lifecycle.
Research scientist role on DeepMind’s AGI Safety and Alignment Team focused on alignment methods, interpretability, AGI control systems, and frontier safety evaluations.
Senior leadership role leading a research team on agentic AI risk modelling and mitigations, focused on advanced AI systems becoming hard to oversee, correct, or shut down and on turning analysis into practical safety interventions.
Research scientist role developing and evaluating white-box methods to improve AI system safety and alignment, with emphasis on realistic evaluations, agentic coding, and threat models.
Technical Program Manager role leading GenAI Safety operations for Gemini and related models, coordinating training, evaluations, red teaming, and safety/alignment initiatives.
Senior manager role leading AI safety delivery, including frontier model evaluations, red teaming, safeguard testing, and client advisory work with government and frontier labs.
Researcher role at ARC focused on AI alignment research, mechanistic understanding of neural networks, and developing training objectives that encourage honest internal reporting.
We are looking for an exceptional writer to help turn promising AI alignment research ideas into compelling, fundable proposals.
The Partnerships Lead at AIAF will be the bridge to the AI safety funding ecosystem, cultivating strong relationships with funders to update them on our work.
Postdoctoral researcher in explainable AI focused on interpreting deep model internals, human-in-the-loop workflows, and rigorous evaluation of interpretability claims; relevant because it explicitly connects interpretability to AI safety and assurance evaluation.
Research scientist role on Gemini Safety and Behavior focused on LLM safety/security, jailbreak mitigation, red and blue teaming, and safety/alignment work for frontier GenAI models.
Penetration tester for Bosch's AI Safety & Security team, running safety evaluations and red-teaming against LLMs and agentic AI systems to find harmful outputs, prompt injection, and tool abuse.
Senior researcher role focused on extreme misuse risks from frontier AI, including technical safeguards, consensus-building, and translating AI safety work into policy-relevant coordination outputs.
Research scientist role on FAR.AI’s Applied White-Box Methods team, developing and evaluating model-internals methods to improve AI safety, alignment, and realistic safety evaluations.
Research Scientist role focused on frontier AI safety and security research, including AI system evaluations, audits, model behavior analysis, robustness, and policy compliance.