Security Engineer – AI (Offensive AI Security and Red Teaming)

Empowering Africa’s tomorrow, together…one story at a time. With over 100 years of rich history and strongly positioned as a local bank with regional and international expertise, a career with our family offers the opportunity to be part of this exciting growth journey, to reset our future and shape our destiny as a proudly African group. Job Summary To provide specialist advice & support in the development & implementation of IT security service delivery processes, methods and techniques enabling secure management & control of IT access, in alignment with governance requirements. Job Description The Security Engineer – with a focus in AI Security and AI Detection & Response (AIDR) is responsible for securing Absa's enterprise Artificial Intelligence (AI), Machine Learning (ML), Large Language Model (LLM), Agentic AI, Retrieval Augmented Generation (RAG), AI Gateway, and Model Context Protocol (MCP) environments. The role combines Security Engineering practices, Security Architecture, Detection Engineering, Threat Hunting, Offensive Security, and Incident Response to protect AI systems from emerging cyber threats while enabling secure innovation across the organisation. It aims to enhance the banks capabilities in AI security controls, AI threat detection, AI red teaming, and AI platform security across cloud, hybrid, and on-premises environments. Key Accountabilities AI Security Architecture Design secure AI and ML architectures. Define AI security standards and reference architectures. Perform AI threat modeling exercises. Conduct security reviews of AI applications and platforms. Establish AI security controls across the AI lifecycle. Secure agentic AI systems and autonomous workflows. Assess security risks associated with AI adoption initiatives. AI Detection & Response (AIDR) Develop AI-specific threat detection use cases. Create monitoring strategies for AI environments. Engineer detections within SIEM, XDR, and EDR platforms. Investigate AI-related security incidents. Develop AI incident response playbooks. Perform AI-focused threat hunting activities. Establish AI attack simulation and validation programs. AI Infrastructure Security Secure enterprise AI ecosystems including: Large Language Models (LLMs) Agentic AI Platforms AI Gateways AI Proxies MCP Servers MCP Clients AI Agents Vector Databases RAG Architectures Model Serving Platforms Model Context Protocol (MCP) Security Develop and implement security controls for MCP ecosystems. Responsibilities include: MCP architecture security reviews Security assessment of MCP servers and connectors Validation of trust boundaries between agents and tools Security testing of MCP integrations Monitoring MCP activity and tool execution Designing secure MCP deployment patterns Developing MCP hardening standards AI Gateway Security Implement and manage security controls for enterprise AI gateways. Responsibilities include: AI traffic inspection Prompt inspection and filtering Content moderation Data Loss Prevention (DLP) Access governance Rate limiting API protection Model routing controls. Offensive AI Security & Red Teaming Conduct offensive security assessments against AI environments. Prompt Security Testing Prompt Injection Indirect Prompt Injection Context Manipulation Jailbreak Testing Prompt Leakage Instruction Override Testing Agent Security Testing Tool Abuse Tool Chaining Attacks Agent Manipulation Agent Privilege Escalation Memory Poisoning Autonomous Agent Abuse RAG Security Testing Knowledge Base Poisoning Retrieval Poisoning Data Exfiltration Vector Database Attacks Unauthorized Data Access Model Security Testing Model Extraction Model Inversion Membership Inference Adversarial Input Testing Model Evasion Model Poisoning Requirements: Bachelor's degree in Computer Science, Cybersecurity, Machine Learning, or a related technical field (or equivalent hands-on experience) Knowledge of offensive security, penetration testing, or red team experience At least 1–2 years focused specifically on AI/ML systems, large language models, or generative AI applications Hands-on experience conducting adversarial testing against LLM-based applications, including prompt injection, jailbreak development, context manipulation, and instruction override techniques Familiarity with published frameworks such as the OWASP Top 10 for LLM Applications, MITRE ATLAS, and NIST AI RMF Working knowledge of agentic AI architectures, including tool use, function calling, ReAct-style loops, and multi-agent orchestration Practical experience identifying tool abuse, tool-chaining attacks, memory poisoning, and privilege escalation paths in autonomous agents Understanding of retrieval-augmented generation (RAG) pipelines, embedding models, and vector databases (e.g., Pinecone, Weaviate, Chroma, FAISS, Milvus) Ability to design and execute knowledge-base poisoning, retrieval poisoning, and exfiltration attacks against RAG systems Familiarity with model-level attacks such as model extraction, model inversion, membership inference, adversarial example generation, evasion, and training-data or fine-tuning poisoning Practical experience with at least one adversarial ML toolkit (e.g., ART, TextAttack, Garak, PyRIT, Counterfit) Proficiency in Python, with the ability to write custom attack scripts, fuzzers, and automation against LLM APIs (OpenAI, Anthropic, Bedrock, Azure OpenAI, self-hosted) Comfort working in Linux environments and with Git Experience with at least one cloud platform (AWS, Azure, or GCP) Solid grounding in application security fundamentals: authentication and authorization flaws, injection classes, SSRF, IDOR, and API security Experience with Burp Suite or equivalent proxy tooling Ability to write clear technical reports translating findings, impact, and remediation guidance for both engineering and executive audiences Ability to run debrief sessions with model developers, ML engineers, and application owners Education Bachelor`s Degrees and Advanced Diplomas: Physical, Mathematical, Computer and Life Sciences (Required) Absa Bank Limited is an equal opportunity, affirmative action employer. In compliance with the Employment Equity Act 55 of 1998, preference will be given to suitable candidates from designated groups whose appointments will contribute towards achievement of equitable demographic representation of our workforce profile and add to the diversity of the Bank. Absa Bank Limited reserves the right not to make an appointment to the post as advertised