Featured
AI Safety Researcher
Security, Risk, Privacy, and Compliance AI Safety and Red Team
£168,000–£252,000/year (£14,000–£21,000/month) — Negotiable
Team and product context
- Team
- Security, Risk, Privacy, and Compliance
- Specialization
- AI Safety and Red Team
- Product family
- Corporate
Security engineering, AI safety, red teaming, privacy, compliance and AI policy.
Explore our products and servicesJob Description
Evaluate AI systems for misuse risk, alignment concerns, reliability gaps, and policy-relevant failure modes. The role produces practical safety findings that engineering and product teams can act on.
Requirements
- Research or applied experience in AI safety, alignment, risk evaluation, or related ML areas
- Familiarity with red-team testing, RLHF review, constitutional AI, or model-risk documentation
- Ability to write clear evidence-backed findings for technical and non-technical readers
- Strong judgement around uncertainty, severity, and mitigation tradeoffs
- Familiarity with red-team testing, RLHF review, constitutional AI, or model-risk documentation
- Ability to write clear evidence-backed findings for technical and non-technical readers
- Strong judgement around uncertainty, severity, and mitigation tradeoffs
Responsibilities
- Design and run safety evaluations for models and AI workflows
- Document findings, mitigations, and residual risks before launch
- Partner with AI red-team, policy, engineering, and product teams
- Improve safety test harnesses and review processes over time
- Document findings, mitigations, and residual risks before launch
- Partner with AI red-team, policy, engineering, and product teams
- Improve safety test harnesses and review processes over time
Benefits
- Direct influence on production AI risk decisions
- Work with technical, policy, and customer-facing teams
- Space to build practical evaluation methods rather than abstract policy only
- Work with technical, policy, and customer-facing teams
- Space to build practical evaluation methods rather than abstract policy only
How we hire
The sequence and format can vary by role.
Application
Initial review
Relevant assessments
Interviews
Decision
Offer and onboarding
Privacy and application documents
Submit only the information and documents requested for this role. For information on personal-data handling, read our Privacy Policy.
Read the Privacy Policy