AI Safety & Security Lead — Threat Modeling

Lilasciences; Greater London; On-site; GBP 144,000 - 230,000; Full time

Benefits offered by this job

Bonus potential

Equity

Comprehensive benefits

Job summary

Lila Sciences is seeking an experienced security practitioner to lead the protection of advanced AI systems at the boundary between AI safety and security. You will investigate real systems, develop threat models, test assumptions, and recommend protections for capabilities and deployments.

You will work with our Security team, researchers and engineers to turn findings into implemented improvements. You will bring security depth and learn model capabilities, evaluations and safeguards through

Qualifications

• Substantial hands-on security experience across multiple disciplines.
• Proven ability to resolve consequential security problems across teams.
• Strong technical judgment in examining unfamiliar systems and code.
• Curiosity about AI safety and scientific systems.
• Clear writing and effective collaboration with specialists.

Responsibilities

• Lead threat modeling and investigate misuses and protections across AI systems.
• Test defenses by examining architectures, code, configurations and evidence.
• Shape protections for safe deployment and convert assessments into concrete requirements.
• Make evidence-based recommendations and plan improvements with risk-based prioritization.
• Collaborate with Security, Safety and engineering teams to implement changes.
• Develop threat-modeling methods and technical guidance for broader use.

Skills

Security architecture

Application security

Infrastructure security

Offensive security

Incident response

Job description

Lila Sciences is seeking an experienced security practitioner to lead the protection of advanced AI systems at the boundary between AI safety and security. You will investigate real systems, develop threat models, test assumptions, and recommend protections for capabilities and deployments.

You will work with our Security team, researchers and engineers to turn findings into implemented improvements. You will bring security depth and learn model capabilities, evaluations and safeguards through