AI Governance Manager in Peterborough

Join Compare the Market and help to make financial decision making a breeze for millions.

Responsibilities

• Provide first-line governance for AI systems across their full lifecycle, from design through to deployment, monitoring, change and incident response.
• Lead technical assessments for AI “License to Operate” decisions, presenting clear, evidence-based recommendations to senior governance forums – including recommendations to elevate, redesign, or withhold approval where warranted.
• Independently evaluate and probe AI systems (including generative and agentic systems) to verify that they are safe, robust, and operating within risk appetite.
• Apply and contribute to evaluation, observability, and safety standards, and translate them into measurable system controls.
• Assess risks across complex AI systems, including multi-component interactions, agentic behaviours, foundation model supply chain risk, and emergent failure modes.
• Produce the technical risk picture for built AI systems – assessment, reporting, and metrics that feed governance forums and CTM's overall AI risk position.
• Consult on the design of high-risk AI systems, feeding in governance requirements early and providing constructive challenge through development.
• Partner with Data Science, ML Engineering, Software Engineering, Risk & Compliance teams, and inform the evolution of governance standards as AI capabilities and practice develop.

Qualifications

• Demonstrable technical depth in AI systems in production – through hands‑on engineering, applied research, or comparable practitioner experience.
• Strong instincts for AI safety and risk: able to identify what could go wrong with a system, reason about likelihood and impact, and recommend mitigations that work in practice.
• Genuine interest in how AI systems fail – including non‑obvious and emergent failure modes – and ability to design and critically assess evaluation approaches for AI systems, including generative AI, RAG, and agentic architectures.
• Comfortable operating as advisor and assessor rather than builder, and comfortable being the dissenting technical voice in the room with senior stakeholders – able to make an evidence-based case for slowing down or rethinking delivery when warranted.
• Clear and concise writing – analysis that equips senior decision-makers to act.
• Background in data science, ML engineering, or AI safety/evaluation research is a strong plus, as is experience with approaches like multi‑step evals, LLM-as‑judge, or red‑teaming, or operating in a regulated consumer or financial services environment.