AI Cybersecurity Engineer – Offensive Security, Threat Modeling

Das ist der Job

Fluency in English.

Darum lohnt es sich

Anticipate sophisticated attacker behaviors to design preemptive security guardrails before models are moved to production. • AI Red Teaming & Offensive Simulations: Conduct targeted offensive operations and ethical hacking against our on-premises AI infrastructure.

Map identified vulnerabilities to ATLAS tactics and techniques to build a robust, AI-specific defensive matrix aligned with traditional MITRE ATT&CK. • On-Premises Infrastructure Hardening: Assess and secure the physical and virtual infrastructure hosting our AI workloads, including bare-metal GPU clusters, local containerized environments (Kubernetes/OpenShift), and private data registries, ensuring strict isolation from external threats. • Proactive Risk Management & Early Detection: Collaborate with the SOC and data engineering teams to build predictive detection use-cases and custom alerts.

Requirements • Professional Experience: 5+ years of dedicated experience in Offensive Security, Red Teaming, or Advanced Penetration Testing, with a proven pivot and at least 2 years of hands-on experience securing or attacking AI/ML workloads. • Banking Sector Experience (Mandatory): Prior experience working within a regulated financial institution in Egypt.

Core Competencies Demonstrates expertise in Offensive Security and Red Teaming, with a strong focus on AI/ML security frameworks and on-premises infrastructure hardening.

Job Responsibilities • Predictive AI Threat Modeling: Lead advanced threat-modeling exercises across the bank’s localized AI/ML pipelines, data repositories, and training environments.

Simulate real-world adversarial attacks, including data poisoning, model inversion, membership inference, evasion attacks, and prompt injection. • Framework Mapping with MITRE ATLAS: Implement and operationalize the MITRE ATLAS (Adversarial Threat Landscape for Artificial-Intelligence Systems) framework.

Identify the early warning signs of an AI-targeted attack or data exfiltration attempt before malicious actors can compromise model integrity.

You must understand the security constraints, compliance mandates, and audit rigor required by the Central Bank of Egypt (CBE) for on-premises operations. • Deep Mastery of AI Attack Frameworks: Expert-level knowledge of MITRE ATLAS and the OWASP Top 10 for Large Language Models (LLMs).

You must have a proven track record of translating these theoretical frameworks into practical, automated offensive playbooks. • On-Prem & Core Technical Skills: Strong proficiency in Python or bash scripting to build custom exploit or testing tools. • Deep understanding of on-premises infrastructure components (Linux administration, private networking, storage area networks). • Direct experience with container security (Docker, Kubernetes) and secure MLOps pipelines. • Education & Certifications: Bachelor’s degree in Computer Science, Cybersecurity, Computer Engineering, or a strictly related field. • Highly valued offensive certifications: OSCP, OSCE, CRTP, or GPEN, alongside any specialized AI/ML security credentials. • Mindset & Communication: A relentlessly curious, adversarial mindset balanced with strong risk management disciplines.

The ability to clearly articulate complex, highly technical AI vulnerabilities into actionable remediation plans for data scientists and senior banking executives. Proficient in risk management and communication of complex vulnerabilities to diverse stakeholders.