Head of AI Safety

This is a fully remote position, open to applicants in Ireland.

📋 Description

• Oversee the delivery, development, and expansion of Moonshot's AI Safety portfolio.

• Direct and ensure quality in applied AI safety initiatives encompassing violence, extremism, CSEA, abuse and grooming, mental health and crisis, as well as child and teen risk categories.

• Provide guidance to frontier AI companies on enhancing model, product, policy, and intervention safety.

• Convert subject-matter expertise into practical recommendations for model safety, policy, product, research, and engineering teams.

• Establish methodological approaches and create structured, testable evaluation frameworks.

• Actively lead and engage in red teaming and adversarial evaluation activities.

• Detect safety failures and formulate suggestions for improving model behavior and user protections.

• Ensure meticulous documentation while adhering to ethical, legal, contractual, data protection, and compliance standards.

• Manage risks related to operations, reputation, delivery, and partnerships.

• Serve as Moonshot's principal applied AI safety representative for AI companies, governments, regulators, and ecosystem partners.

• Cultivate relationships with technical teams, governments, foundations, regulators, academics, researchers, civil society organizations, and practitioners.

• Represent Moonshot in external meetings, briefings, workshops, and sector engagements.

• Lead, mentor, and manage the AI safety team.

• Assist in workforce planning, performance management, professional development, and team well-being.

• Collaborate with operations, finance, research, and technical teams.

• Expand the AI safety portfolio through strategic opportunities, partnerships, and funding sources.

• Spearhead proposal development, scoping, renewals, and business development efforts.

• Create repeatable methodologies, service offerings, and partnerships.

• Aid in communications, publications, briefings, and thought leadership initiatives.

• Supervise project planning, staffing, budgeting, forecasting, and delivery timelines.

⛳️ Requirements

• Proven experience in trust & safety, online harms, violence prevention, safeguarding, or public health, with the ability to adapt knowledge to AI systems.

• A strong curiosity about AI and the capacity to quickly develop technical fluency.

• Experience in designing research, evaluation frameworks, or interventions related to violent extremism, CSEA, self-harm and crisis, or targeted violence.

• Demonstrated ability to manage projects, teams, budgets, partners, and clients effectively.

• Strong skills in people management.

• Exceptional written communication skills tailored for government, foundation, or enterprise audiences.

• Resilience and comfort in handling highly sensitive or graphic content.

• Awareness of well-being practices when dealing with sensitive content.

• Sound judgment and the ability to navigate ambiguity, competing priorities, and sensitive stakeholder environments.

• Willingness to travel and work outside regular hours when necessary.

• Trustworthiness, discretion, diplomacy, and readiness to undergo relevant security clearance procedures.

• Experience in supporting business development, grant funding, or procurement.

• Strong commitment to Moonshot's mission.

• Eligibility to work in the UK.

• Ability to pass necessary security clearance procedures.

• Direct experience in model safety, red teaming, or adversarial evaluation of LLMs or other AI systems (desirable).

• Understanding of LLM architecture, safety tooling, or trust & safety policy (desirable).

• Experience in child safety evaluation, teen-safety product work, or grooming and CSEA detection (desirable).

• Experience in government or regulatory engagement (desirable).

• Experience in designing intervention or diversion programs (desirable).

• Academic or applied background in radicalisation studies, forensic psychology, or violence risk assessment (desirable).

• Familiarity with taxonomy or classifier development and testing data workflows (desirable).

🏝️ Benefits

• 30 days of paid annual leave, excluding public holidays.

• Flexible public holiday policy allowing the option to work public holidays in exchange for a day off at another time.

• Private healthcare package that includes access to specialist mental health coverage for partners and children.

• Dental and vision insurance.

• Life insurance and income protection.

• Employee Assistance Programme providing access to mental health support.

• 26 weeks of paid maternity leave.

• 8 weeks of paid paternity leave.

• Share options granted to all permanent employees upon employment.