VP-AI Audit
Mashreq · India
Key Result Areas
Develop and implement risk-based AI/GenAI audit strategies aligned with the Bank's AI agenda and regulatory expectations.
Execute audits over the AI/ML lifecycle — data sourcing, training, validation, deployment, monitoring, retraining, and decommissioning (MLOps/LLMOps).
Provide assurance on AI governance, model risk management (MRM), ethics, fairness, bias, explainability, and human-in-the-loop controls.
Audit GenAI/LLM use cases — RAG pipelines, fine-tuning, prompt engineering, guardrails, vector databases, and output validation.
Assess AI cybersecurity risks — adversarial attacks, prompt injection, data poisoning, model theft, jailbreaks (OWASP LLM Top 10, MITRE ATLAS).
Evaluate third-party AI risks covering foundation model providers (OpenAI, Anthropic, Google, Meta, Mistral, open-source) and cloud AI platforms (Azure OpenAI, AWS Bedrock, Google Vertex AI).
Assess compliance with AI and data regulations — CBUAE, QCB, SBP, RBI,UAE PDPL etc.
Audit AI use in credit, AML/fraud, KYC, chatbots, personalization, trading, and operations automation.
Prepare and present impactful audit reports to the Board Audit Committee, GCEO, and senior management, translating complex AI concepts into business language.
Partner in Internal Audit AI transformation — continuous auditing, GenAI-enabled audit tools, and audit team upskilling.
Guide, coach, and develop AI audit team members; foster a culture of learning, agility, and innovation.
Support integrated audits by providing AI/technology subject-matter expertise across the Bank.
Knowledge, Skills and Experience
Education
Bachelor's degree in Computer Science, IT, Data Science, AI, Statistics, Mathematics, or a related quantitative field; Master's in AI/ML or Data Science preferred.
Experience
Minimum 10–12 years in IT audit, technology risk, model risk, or AI/data governance, with at least 3–4 years directly focused on AI/ML or GenAI risk, governance, or audit, preferably in banking.
Certifications
CISA mandatory (or to be obtained within 12 months).
One or more preferred: ISACA AAIA (Advanced in AI Audit), CISSP, CRISC, CGEIT, CDPSE.
Technical Knowledge
Strong understanding of AI/ML concepts — supervised, unsupervised, reinforcement, deep learning, NLP, computer vision.
GenAI and LLMs — foundation models, transformers, embeddings, RAG, fine-tuning (SFT, RLHF, LoRA), prompt engineering, agentic and multi-modal AI.
Familiarity with major model versions and providers — OpenAI (GPT-4/4o/5), Anthropic (Claude), Google (Gemini), Meta (Llama), Mistral, and leading open-source models.
AI platforms/tooling — Azure OpenAI, AWS Bedrock/SageMaker, Google Vertex AI, Databricks, Hugging Face, LangChain, vector databases.
AI governance and risk frameworks
Skills
Strong analytical and problem-solving skills focused on novel AI risks.
Excellent communication and interpersonal skills to convey complex AI concepts to technical and non-technical stakeholders, including the Board.
Ability to work independently, lead a team, and collaborate across departments and geographies.
Added Advantages
Hands-on involvement in any part of an organization's AI initiatives (use case build, model validation, AI governance council, MLOps, GenAI product).
Banking / financial services domain knowledge (credit, fraud/AML, digital channels, compliance).
Experience with AI-enabled internal audit tools and audit analytics.