AI SECURITY ENGINEER

CARE International; Atlanta (GA); On-site; USD 120,000 - 160,000; Full time

Job summary

CARE International in Atlanta, GA seeks an AI Security Engineer to safeguard CARE's AI systems, ML pipelines, and agentic AI deployments, ensuring secure adoption aligned with humanitarian mission.

The role leads threat modeling, red-team testing, and incident response, partnering with IT, data, legal, and program teams to embed governance, privacy, and security into the AI lifecycle.

Qualifications

• Experience securing AI/ML systems and threat modeling.
• Knowledge of NIST AI RMF, ISO/IEC 42001, and AI security governance.
• Ability to partner with IT, legal, and program teams to align security with mission.

Responsibilities

• Assess and harden AI/ML models, APIs, and inference endpoints against adversarial attacks.
• Design and implement security controls across the AI development lifecycle with least-privilege access.
• Develop AI security policies and risk frameworks aligned with industry standards and audits.
• Lead incident response for AI-related security events and deliver security awareness on AI threats.

Skills

Threat modeling

Red-team testing

Incident response

Governance & compliance

Security architecture

AI security

Tools

Microsoft 365 Copilot

Anthropic Claude

Copilot Studio

Graph-grounded agents

MCP connectors

Job description

JOB SUMMARY:

The AI Security Engineer safeguards CARE's AI systems, machine learning pipelines, and agentic AI deployments against emerging threats — including enterprise platforms such as Microsoft 365 Copilot and Anthropic Claude. This role embeds security into the AI development lifecycle, establishes governance frameworks aligned with NIST AI RMF and ISO/IEC 42001, and leads threat modeling, red-team testing, and incident response for AI-related risks. The position collaborates cross-functionally with IT, data, legal, and program teams to ensure responsible, secure AI adoption in support of CARE's humanitarian mission.

RESPONSIBILITIES:

AI System Security & Threat Assessment

• Assess and harden AI/ML models, APIs, and inference endpoints against adversarial attacks including prompt injection, model inversion, and data poisoning. Perform threat modeling for generative AI and agentic AI deployments. Conduct red-team exercises against LLMs and AI-powered applications, including jailbreak testing and adversarial prompt analysis. Monitor AI workloads for anomalous behavior, model drift, and signs of compromise. This responsibility directly protects CARE's operational data and the integrity of AI-assisted program delivery.

Security Architecture & Enterprise AI Integration

• Partner with engineering and IT teams to integrate security controls into the AI/ML development lifecycle. Design and enforce least-privilege access controls for AI systems and agents. Evaluate and secure enterprise AI platforms — including Microsoft 365 Copilot (plugins, Copilot Studio, Graph-grounded agents) and Anthropic Claude (API integrations, Claude for Enterprise, MCP connectors) — assessing data access scope, permission boundaries, and supply chain exposure. Establish secure data handling practices for training data and model artifacts.

AI Governance & Compliance

• Develop and maintain AI security policies, standards, and risk frameworks aligned with NIST AI RMF, ISO/IEC 42001, OWASP LLM Top 10, and MITRE ATLAS. Maintain an inventory of AI systems and assess risk posture, access scope, and data exposure. Support audits and compliance reviews related to AI use. Advise senior leadership (CISO, COO, General Counsel) on evolving AI security risks, threat intelligence, and remediation priorities. Ensure AI adoption remains compliant with applicable data privacy laws and organizational policy.

Incident Response & Security Awareness

• Lead or support incident response for AI-related security events including prompt injection attacks, model abuse, and data exfiltration through AI interfaces. Develop playbooks and runbooks specific to AI security incidents. Deliver security awareness training on AI-generated threats such as deepfakes, AI-assisted phishing, and social engineering. Stay current on AI security research, threat actor tactics, and responsible disclosure in the AI security community.