AI Safety & AI Lead

Jersey City, New Jersey, United States (Onsite)

lGovernance Lead / Senior Manager or Director-level Specialis

se Define and operationalize Responsible AI practices across the AI lifecycle for AIRP and enterprise citizen-development initiatives. The role ensures AI systems are safe, fair, explainable, transparent, compliant, monitored, and aligned with enterprise values, model risk, legal, compliance, data governance, cybersecurity, and audit expectation

• isThe organization is aiming to democratize AI responsibly; this role must support enterprise AI plus citizen development through Microsoft Power Platform, Copilot Studio, Power Apps, Power Automate, and Power B
• I.Governance must be practical enough to support business AI use cases while satisfying banking, model risk, security, privacy, and audit control
• s.The candidate should be able to govern high-risk workflows such as KYC, credit underwriting, financial crime, and sanctions screenin

• ip Responsible AI policy, control framework, risk taxonomy, governance workflows, and production-readiness criteria for AIRP and citizen AI use case
• s.AI risk assessments, impact assessments, safety evaluations, model-risk alignment, and post-production monitoring standard
• s.Cross-functional alignment across engineering, product, legal, compliance, model risk, audit, cybersecurity, data governance, and citizen-development enablement team

• es Define Responsible AI standards, policies, procedures, risk-classification methods, and operating models for AI and GenAI initiative
• s.Establish governance processes for use-case intake, risk assessment, model review, approval workflows, deployment readiness, ongoing monitoring, and issue escalatio
• n.Develop safety and evaluation frameworks covering fairness, bias, explainability, transparency, robustness, privacy, hallucination, harmful outputs, human oversight, and overrelianc
• e.Define guardrail requirements for LLMs, RAG systems, agentic workflows, high-risk banking applications, and citizen-development solution
• s.Partner with model risk, legal, compliance, data governance, cybersecurity, audit, product, engineering, and business teams to align AI controls with enterprise expectation
• s.Lead AI impact assessments, risk reviews, control assessments, readiness reviews, remediation planning, and AI incident escalation processe
• s.Establish metrics and monitoring for bias indicators, safety violations, explainability gaps, harmful outputs, hallucination trends, user feedback, and behavior drif
• t.Create governance playbooks and reusable control evidence for AIRP use cases and Power Platform / Copilot Studio citizen-development workflow

• leDeep understanding of Responsible AI, AI ethics, model governance, model risk, explainability, fairness, privacy, safety, and enterprise risk managemen
• t.Experience implementing AI governance or Responsible AI controls in production or enterprise environment
• s.Understanding of LLM-specific risks such as hallucination, bias, toxicity, prompt injection, data leakage, over reliance, unsafe automation, and human oversight gap
• s.Ability to translate policy and regulatory expectations into practical product, engineering, operating, and audit control
• s.Experience working with cross-functional risk, compliance, legal, security, data, audit, product, and engineering stakeholder
• s.Ability to define controls that scale across centralized AI platforms and distributed citizen-development adoptio

• ce Experience in banking, insurance, fintech, consulting, regulatory risk, model risk management, technology governance, or data governanc
• e.Experience building AI risk taxonomies, control libraries, governance operating models, Responsible AI playbooks, or model-risk-aligned review processe
• s.Familiarity with Power Platform, Microsoft Copilot Studio, Power Apps, Power Automate, Power BI, global AI governance frameworks, model validation practices, privacy regulation, and audit expectation
• nsWhat Responsible AI framework have you implemented, and how was it operationalize
• d?
  
  How do you classify AI use-case risk in a regulated enterpris
• e?
  
  How would you govern KYC, credit underwriting, financial crime, or sanctions screening AI use case
• s?
  
  How do you govern citizen development through Copilot Studio, Power Apps, Power Automate, and Power B
• I?
  
  How do you evaluate and monitor hallucination, bias, fairness, explainability, and human oversigh
• t?
  
  How do you balance innovation speed with control expectation