AI Risk Oversight, Director

BlackRock•New York, NY

5d•$210,000 - $270,000•Hybrid

About The Position

BlackRock is seeking a Director to lead AI Risk Oversight within its Enterprise Risk Management function, part of Risk & Quantitative Analysis (RQA). RQA provides independent oversight of the firm’s fiduciary and enterprise risks. ERM focuses on risks affecting the corporate entity, including operational, technology, third-party, model, and data risk. As a second line of defense, the function ensures that senior management defines and implements effective controls to protect clients and the firm while supporting business goals within the firm’s risk tolerance. The role provides independent oversight and challenge across the AI lifecycle, from how AI systems are designed and governed to how they access data and systems, operate, and enter production. It addresses risks spanning AI models, including agentic AI, deployment, infrastructure and architecture, access management, third-party services, cybersecurity dependencies, production monitoring, and AI compute and inference. The Director partners with Model Risk Management, Technology and Cyber Risk, Third Party Risk, Information Security, Engineering, and first-line AI owners to develop an integrated view of AI risk. The role calls for technical understanding of modern AI systems and the risk and governance frameworks needed to deploy them safely at enterprise scale. The Director will build and lead an AI risk capability within RQA, challenge senior AI, engineering, and business leaders, and communicate complex risks to executive and board-level audiences. The position is based on a hybrid work model: employees are currently required to work in the office at least four days per week, with flexibility to work from home one day per week. Some business groups may require additional office time. For New York, NY, the stated salary range is USD $210,000–$270,000. Employees are also eligible for an annual discretionary bonus and benefits including healthcare, leave, and retirement benefits. Total compensation may vary based on role, location, and firm, department, and individual performance.

Requirements

Enterprise experience in model risk, AI/ML governance, technology risk, enterprise risk, engineering risk, model implementation, or a closely related control discipline.
Strong understanding of the AI/ML lifecycle, including model development, validation, integration, deployment, inference, monitoring, and production operations.
Experience with model governance and model implementation, and ability to distinguish traditional model validation from broader technology and operational risks of enterprise AI deployment.
Strong understanding of modern AI architecture and infrastructure, including model APIs, data and inference pipelines, cloud environments, compute/GPU infrastructure, access management, and production deployment patterns.
Understanding of generative and agentic AI risks, including hallucination, reliability, data access, third-party models, and AI use in automated or agentic workflows.
Experience assessing third-party technology or AI risk, including dependencies on external models, platforms, cloud infrastructure, and AI service providers.
Ability to assess technical and operational control effectiveness and provide credible independent challenge to senior engineers, AI practitioners, and business leaders.
Demonstrated experience developing risk metrics, indicators, and senior management reporting that translates technical issues into clear risk assessments.
Excellent judgment and critical reasoning, including the ability to distinguish theoretical AI risks from material risks in real-world enterprise production environments.
Strong communication and influencing skills, with the ability to engage credibly with senior technology, risk, and business executives.
Bachelor’s degree in computer science, engineering, mathematics, statistics, finance, risk management, or a related discipline is required.

Nice To Haves

Experience in financial services or another highly regulated environment is preferred.
Familiarity with established and emerging AI and technology risk frameworks, such as the NIST AI Risk Management Framework and NIST Cybersecurity Framework, and with model risk management principles.
An advanced degree or relevant professional certification.

Responsibilities

Provide independent oversight, review, and challenge of AI risk across the firm, assessing design, governance, implementation, deployment, and monitoring, and holding first-line owners accountable for managing risk within the firm’s risk tolerance.
Assess AI technology architecture and operating environments, including model and data pipelines, access and entitlements, production deployment, cloud and compute infrastructure, GPU utilization, and cybersecurity dependencies.
Oversee production risks arising from AI models, including human oversight and controls for generative and agentic AI, in partnership with Model Risk Management.
Partner with Third Party Risk to assess end-to-end risks from external AI models, platforms, and infrastructure, including vendor dependencies, data and system access, model changes, cloud and compute concentration, and controls from selection through production deployment.
Provide independent oversight of AI-related incidents and control issues; challenge root-cause analysis and remediation, identify systemic or emerging themes, and ensure material risks and unresolved issues are escalated appropriately.
Develop an enterprise view of AI risk using metrics, risk indicators, and senior-level reporting that translate technical risks into clear assessments of residual risk for senior management, committees, and boards.
Connect AI risk work across Model Risk Management, Technology and Cyber Risk, Third Party Risk, Data Risk, Legal and Compliance, Privacy, and business technology teams to promote clear accountability and integrated risk management.
Build and lead an AI risk capability within RQA, challenge senior AI, engineering, and business leaders, and communicate complex AI risks to executive- and board-level audiences.