AI Security Engineer / Offensive Security Engineer - Technology

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a AI Security Engineer / Offensive Security Engineer - Technology based in Brazil.

As an AI Security Engineer, you will continuously test and strengthen AI-powered products, infrastructure, APIs, and agentic systems.
You will take a hands-on offensive security approach, actively identifying vulnerabilities and reproducing exploits before partnering with engineers on remediation.
The role covers emerging AI-specific risks such as prompt injection, tool misuse, data leakage, permission escalation, memory poisoning, and unsafe agent behavior.
You will combine security expertise with engineering skills to build automation, testing, monitoring, and defensive capabilities.
Working in a fast-moving environment, you will help evolve security practices alongside a rapidly developing product.
This is an opportunity to make security an ongoing engineering discipline while protecting sensitive identities, permissions, data, and user workflows.

• Continuously red-team products, infrastructure, APIs, and agentic systems to identify security weaknesses and meaningful attack paths.
• Discover, reproduce, debug, and validate vulnerabilities, then work directly with engineering teams through remediation.
• Test AI and LLM-specific attack surfaces, including prompt injection, tool misuse, data leakage, permission escalation, memory poisoning, and unsafe agent behavior.
• Use AI agents, automation, and security tooling to continuously probe systems, accelerate vulnerability discovery, and support triage.
• Prioritize security findings based on exploitability, user impact, and overall business risk.
• Strengthen authentication, authorization, identity management, secrets management, data boundaries, and cloud infrastructure security.
• Build lightweight monitoring, testing, and defensive systems that continuously improve as the product and threat landscape evolve.
• Establish pragmatic security standards and development practices while avoiding unnecessary processes and complexity.

Requirements:

• Strong professional experience in application security, product security, cloud security, offensive security, or a related discipline.
• Hands-on ability to identify, exploit, debug, and remediate vulnerabilities.
• Solid understanding of modern web architectures, APIs, authentication and authorization, cloud infrastructure, and data security.
• Strong software engineering fundamentals and sufficient coding skills to develop security tooling, automation, and remediation solutions.
• Familiarity with AI and LLM security risks, with a strong interest in securing agentic systems.
• Practical understanding of established security best practices and the judgment to determine when deeper investigation is warranted.
• Comfortable working in a fast-moving, zero-to-one environment where security practices and threat models evolve alongside the product.
• Experience red-teaming LLM or agent systems, building automated security testing infrastructure, or securing products involving sensitive identity, memory, permissions, or user data is highly valuable.
• Fluent English communication skills.
• Located in Latin America, with Brazil as the target location for this opportunity.

Benefits:

• 100% remote work with the flexibility to work from a location that supports productivity.
• Highly competitive compensation paid in USD.
• Paid time off to support rest and work-life balance.
• Flexible, autonomous work environment focused on results rather than hours.
• Opportunity to work on innovative, high-impact projects for leading U.S. companies.
• Access to a diverse global network of 600+ technology professionals across 25+ countries.
• Collaboration with experienced and senior technology professionals.
• Engagement activities and a culture focused on employee well-being and professional development.
• Exposure to cutting-edge AI security, agentic systems, cloud infrastructure, and offensive security challenges.