GDS Cyber - Frontier AI Layered Defense - Staff Consultant

At EY, you’ll have the chance to build a career as unique as you are, with the global scale, support, inclusive culture and technology to become the best version of you. And we’re counting on your unique voice and perspective to help EY become even better, too. Join us and build an exceptional experience for yourself, and a better working world for all. 

Job Title: Staff / Junior AI Security Engineer – Frontier AI, Layered Defense and AI Security – Technology Consulting

EY is seeking a Staff / Junior AI Security Engineer to join the Frontier AI and Layered Defense practice. This role is designed for an early-career professional who will support the secure design, implementation, testing, and monitoring of AI-enabled systems, including LLM applications, RAG pipelines, and emerging agentic AI workflows.

The position will focus on applying layered defense principles across the AI lifecycle, including secure data access, prompt and context controls, model interaction safeguards, tool-use boundaries, response quality checks, monitoring, and appropriate escalation paths. The candidate will contribute to practical implementation activities while learning how frontier AI risks such as unintended information exposure, unreliable responses, unsafe tool invocation, misuse patterns, and data leakage are addressed in enterprise environments.

This is an execution-focused role that will work under the guidance of senior AI security engineers, architects, and cybersecurity teams. The successful candidate will help configure application safeguards, execute predefined validation scenarios, validate RAG and agent workflows, review logs and telemetry, document findings, and support secure integration of AI applications with enterprise platforms and APIs.

The ideal candidate has foundational knowledge of software development, cybersecurity, data handling, or AI/ML, along with strong curiosity about frontier AI security. The role does not require deep AI security specialization at entry; however, the candidate should be able to learn quickly, follow structured security playbooks, apply secure coding practices, and collaborate effectively in delivery teams.

This role provides hands-on exposure to modern AI security engineering, including application safeguard implementation, secure RAG patterns, agent workflow control checks, observability, and reusable automation that strengthens enterprise AI security capabilities.

Key Responsibilities:

• Support implementation of layered defense controls for LLM, RAG, and agentic AI use cases, including input handling, context isolation, tool-use boundaries, response checks, access controls, and monitoring.
• Assist in building and testing AI applications using frameworks such as LangChain, LangGraph, LlamaIndex, AutoGen, OpenAI-compatible APIs, and related orchestration tools.
• Configure and validate basic AI application safeguards, including prompt handling, response constraints, sensitive data handling checks, and escalation paths for uncertain or high-impact outputs.
• Support secure RAG implementation by helping validate data ingestion, retrieval boundaries, embedding and vector store access, source attribution, and secure handling of structured and unstructured enterprise data.
• Execute predefined misuse-resistance and scenario validation checks, including attempts to bypass instructions, expose hidden context, trigger unintended actions, or produce unsafe or unreliable outputs.
• Review AI system logs, traces, prompts, outputs, tool calls, and telemetry to identify anomalies, unexpected behavior, and potential security issues for escalation.
• Support secure integration of AI systems with enterprise APIs, identity platforms, cloud services, workflow tools, and knowledge repositories under senior guidance.
• Assist in documenting validation results, control observations, implementation notes, remediation actions, and reusable delivery patterns.
• Contribute to automation scripts, test harnesses, and repeatable playbooks for AI application validation and continuous monitoring.
• Follow secure coding practices, data protection requirements, internal standards, and responsible technology expectations while working on AI applications and integrations.
• Stay current on emerging frontier AI risks, AI application security patterns, resilience testing methods, and layered defense practices, and apply learnings to project delivery.

Technical Skills and Expertise:

• 0–3 years of experience in software development, cybersecurity, AI/ML, data engineering, cloud engineering, or related academic/project work.
• Foundational understanding of AI/ML concepts, including LLMs, prompts, embeddings, tokens, vector databases, RAG, and basic agent workflows.
• Familiarity with Python and basic scripting for automation, testing, data processing, or API integration.
• Working knowledge of SQL and basic data handling concepts, including structured and unstructured data sources.
• Awareness of AI application risks such as unintended information exposure, data leakage, unreliable outputs, unsafe tool use, insecure integrations, model misuse, and over-permissive automation.
• Foundational knowledge of cybersecurity concepts including authentication, authorization, IAM, API security, secrets handling, secure coding, logging, and vulnerability management.
• Exposure to cloud environments such as Azure, AWS, or GCP, with basic understanding of secure deployment and access configuration.
• Familiarity with AI or application development frameworks such as LangChain, LangGraph, LlamaIndex, AutoGen, OpenAI APIs, or comparable tools is preferred.
• Basic understanding of CI/CD pipelines, version control, software testing, and secure software development lifecycle practices.
• Ability to follow structured validation plans, implement predefined controls, document observations, and escalate risks clearly.
• Strong communication skills, attention to detail, and willingness to learn in a fast-evolving frontier AI security domain.

Preferred Skills and Exposure

• Internship, academic, or project experience involving AI/ML, LLM applications, chatbots, RAG solutions, or agent-based workflows.
• Exposure to AI application security or broader application security concepts such as LLM-specific risk patterns, API gateways, IAM, secrets management, endpoint security, or secure SDLC.
• Basic understanding of secure data pipelines, retrieval architectures, vector databases, enterprise search, or API-based integrations.
• Familiarity with prompt design, prompt evaluation, LLM output checks, application safeguard configuration, or AI quality evaluation.
• Exposure to testing tools, automation scripts, notebooks, logging platforms, or observability dashboards used for AI or application behavior analysis.
• Interest in frontier AI, agentic AI workflows, secure AI application patterns, and defense-in-depth approaches for enterprise AI systems.
• Experience working in Agile, collaborative engineering, cybersecurity, or consulting delivery environments.

EY | Building a better working world 

 
EY exists to build a better working world, helping to create long-term value for clients, people and society and build trust in the capital markets.  

 
Enabled by data and technology, diverse EY teams in over 150 countries provide trust through assurance and help clients grow, transform and operate.  

 
Working across assurance, consulting, law, strategy, tax and transactions, EY teams ask better questions to find new answers for the complex issues facing our world today.