Manager of AI Governance, Dow Jones

• Typically 4 - 7 years of relevant experience in AI or data engineering, technical AI governance, security architecture, technology risk, model validation or a related discipline, including practical experience assessing or deploying AI systems
• Strong understanding of LLMs, retrieval-augmented generation, AI agents and tool calling, including their limitations and failure modes. Familiarity with traditional machine learning and model evaluation is also valuable
• Ability to review architecture diagrams, API documentation, access policies, configuration files and logs, and to discuss implementation choices confidently with engineers
• Practical proficiency in Python or a comparable scripting language sufficient to inspect integrations, query APIs, analyse results and prototype control checks or evaluations
• Working knowledge of cloud architectures, identity and access management, authentication, secrets management, application security and data protection
• Experience translating risk or policy requirements into implementable controls and assessing evidence of their effectiveness
• Strong judgement in distinguishing material risks from routine implementation choices and recommending controls proportionate to the use case
• Ability to explain technical findings clearly to business leaders, editorial stakeholders and executive committees, and to work constructively across teams
• A degree in Computer Science, Engineering, Data Science, Information Security or a related discipline, or equivalent practical experience

Desirable

• Reporting to the VP, Data & AI Governance, you will lead the technical assessment and assurance of AI systems across Dow Jones, including internally developed solutions, third-party tools and AI capabilities embedded in existing platforms
• You will examine how AI systems are designed, what data they access, which actions they can perform and how their risks are controlled. As adoption expands into agentic AI, you will help establish the requirements for systems that operate on our behalf, including those that modify records, execute transactions or interact with enterprise applications
• Working closely with Engineering, Architecture, Cyber Security, Legal, Privacy and business owners, you will turn governance requirements into controls that teams can implement and demonstrate. You will also develop reusable assessment and approval approaches that allow teams to move quickly when established requirements are met
• Lead technical AI risk assessments. Review architectures, data flows, model configurations, integrations and deployment environments. Identify risks, recommend proportionate mitigation and document residual risks and approval conditions using the Dow Jones AI Risk Matrix
• Assess agentic capabilities and system access. Establish which tools and actions an agent can use, under whose identity and permissions, and with what degree of autonomy. Evaluate access through MCP, APIs, command-line interfaces and browser automation, including potential changes to enterprise systems, human approval requirements, action limits and recovery mechanisms
• Evaluate changes to previously approved tools. Determine whether AI or agent access introduces new users, permissions, data flows, scale of activity or autonomy beyond the original approval. Reuse existing assessments and focus additional review on material changes
• Define and validate technical controls. Partner with technical owners to establish least-privilege access, credential protection, data handling restrictions, human oversight, audit logging and mechanisms to suspend or revoke access. Review implementation evidence and test results to establish whether controls work as intended
• Assess AI-specific security and data risks. Work with Cyber Security and Privacy to address prompt injection, unintended disclosure, inappropriate tool use and risks introduced by external models, connectors and dependencies. Evaluate how sensitive information, proprietary content and personal data enter model context, retrieval systems, outputs and logs
• Establish AI evaluation and monitoring requirements. Define proportionate tests for accuracy, groundedness, reliability, bias and safe behaviour, including tool selection and execution for agents. Work with delivery teams to set acceptance criteria and monitor failures, control effectiveness and changes in system behaviour
• Build scalable governance processes. Develop technical standards, assessment templates and approved integration patterns. Establish clear routes for routine approvals, exceptions and reassessment when models, tools, permissions or intended uses change
• Conduct technical vendor assessments and support pilots. Examine vendor architecture, data handling, access controls, evaluation evidence and monitoring capabilities. Define risk and performance criteria for proofs of concept and assess readiness for production use
• Support governance throughout the AI life cycle. Maintain visibility into system owners, risk assessments, dependencies, approval conditions and outstanding mitigation. Contribute to incident reviews and ensure lessons inform future controls and assessments
• Provide clear technical advice. Present findings and recommendations to the AI Steering Committee and other decision-makers. Translate policy and regulatory requirements into technical guidance with Legal and Privacy, and contribute targeted training for teams building or deploying AI