Jobs at FAR.AI

Conduct applied AI safety research using white-box methods such as activation explainers, steering, patching, attribution, and probes. Develop, evaluate, and stress-test methods on realistic large-scale models and long-context agentic coding tasks. Own research directions, publish findings, communicate results to technical and non-technical audiences, and collaborate with frontier labs, governments, and academics.

Own full-cycle accounting for a growing nonprofit, including month-end and year-end close, journal entries, reconciliations, general ledger maintenance, grant and fund accounting, financial reporting, audit coordination, and development of scalable controls and processes. The role partners with finance leadership and program teams while supporting U.S. and international operations, including a Singapore entity.

Operate and scale Kubernetes GPU clusters supporting frontier AI research. Responsibilities include node lifecycle management, batch scheduling, multi-tenancy, storage, fault-tolerant distributed training, security hardening, infrastructure as code, observability, provider integration, and direct infrastructure support for research teams. The role requires production Linux and Kubernetes experience, programming ability, and ownership of large-scale GPU, HPC, or batch infrastructure.

Lead the GPU cluster infrastructure team while remaining at least 50% hands-on. Own technical direction, architecture, scheduling, storage, security, infrastructure as code, observability, incident response, and platform roadmap. Manage hiring, coaching, project prioritization, and provider and researcher escalations. Operate production Kubernetes GPU clusters across multiple providers and improve performance, reliability, fault tolerance, and workload isolation for large-scale AI research.

Own FAR.AI’s internal IT and security function, including accounts, devices, identity and access management, onboarding, offboarding, office networking, support, endpoint hardening, data protection, governance, and vendor risk. The role will remediate security gaps using threat modeling, automate self-service infrastructure, partner across teams, communicate and manage change, and hire and lead an IT Coordinator.

Lead FAR.AI’s pre-training safety research agenda, focusing on capability control through data filtering, gradient routing, unlearning, and synthetic data. Build and mentor a technical team, direct large-scale experiments on frontier language models, collaborate with red teams, evaluate safety-capability tradeoffs, publish findings, and drive adoption through partnerships, grantmaking, events, and policy engagement. The role requires hands-on research and engineering alongside strategic leadership.

Leads engineering for FAR.AI’s AI red team, combining hands-on development with team management. Builds scalable red-teaming tools, agentic attack systems, evaluation frameworks, vulnerability databases, and reporting infrastructure for frontier AI models. Manages hiring, mentoring, sprint planning, engineering execution, and high-stakes engagements with AI companies and governments. Partners on technical and product strategy while translating red-team findings into improved safeguards and broader AI safety impact.

As Jailbreaking Lead, you will engineer jailbreaks for frontier AI models, lead red-teaming engagements, mentor team members, and shape the research agenda to address vulnerabilities in AI systems.

Lead content and strategy for a 12+ event annual portfolio: design agendas, curate speakers and attendees, own pre-event framing and speaker engagement, produce written outputs, steward post-event continuity and community building, manage stakeholder relationships across AI safety, and collaborate with internal teams to translate event visions into participant-facing programs and systems.

Lead end-to-end project planning and delivery for multiple concurrent events, managing logistics, vendors, budgets, on-site execution, and post-event wrap. Coordinate cross-functional dependencies, enable junior PMs, build delivery systems, surface risks, and make real-time resourcing decisions. Travel up to 40%.

Lead end-to-end technical programs for AI safety research and infrastructure: scope plans, prioritize across teams, build compute/resource roadmaps, enable technology transfer, and coordinate internal stakeholders to accelerate research delivery.

Support the Business Operations Manager and COO by owning operational tickets, compliance tracking, finance and HR operations, IT coordination, FAR.Labs logistics, and ad-hoc projects. Triage requests, maintain policies and calendars, coordinate vendors and equipment, and improve processes to keep the organization running smoothly.

The Senior Research Engineer will lead projects in AI safety, focusing on machine learning, high-performance computing, or technical leadership. Responsibilities include detecting model deception, preventing misuse, and developing frameworks for research acceleration.

The Research Lead will develop and lead a research agenda on AI safety, manage a team, and share findings through publications and presentations.

As a Research Engineer at FAR.AI, you will develop machine learning algorithms, run experiments, collaborate with researchers, and contribute to research papers.

The Research Scientist will lead and develop AI Alignment research agendas, publish findings, and contribute to best practices in AI safety research.

FAR.AI invites interested candidates to express their interest for future roles related to AI safety research and advancement.