AI Security Engineer

Truist•Charlotte, NC

2d•Onsite

About The Position

The AI Security Engineer helps design, implement, test, and operate the controls that keep enterprise AI systems safe, governed, and production ready. This role focuses on the security engineering foundations required for AI-enabled applications, agents, prompt-driven workflows, and tool-integrated automations operating in a regulated enterprise environment. This is a hands-on engineering role within the Forge AI Security & Governance model. The engineer supports guardrail implementation, prompt-injection defense, output filtering, monitoring, secure tool-use boundaries, logging, detection content, and deployment-readiness controls for AI-enabled systems. The work spans design, testing, automation, detection engineering, and operational support across the AI delivery lifecycle. Daily work includes implementing security controls for AI and agentic systems, validating configurations, supporting adversarial test preparation, building monitoring logic, partnering with engineering to harden prompt and tool behaviors, documenting controls, and ensuring AI solutions meet enterprise safety, traceability, and governance requirements before and after deployment.

Requirements

Bachelor’s degree and five years of experience in systems engineering or administration or an equivalent combination of education and work experience
In-depth knowledge in applied enterprise information security technologies including but not limited to firewalls, intrusion detection/prevention systems, network operating systems, identity management, database activity monitoring, encryption, content filtering, and Mainframe security
Previous experience in planning and managing IT projects
3+ years of experience in security engineering, cybersecurity operations, application security, or a closely related technical discipline.
Hands-on experience implementing technical controls for enterprise software, APIs, cloud-native services, or automation workflows.
Working knowledge of AI/LLM security concepts such as prompt injection, unsafe output handling, tool-use abuse, sensitive data exposure, and control boundary enforcement.
Experience with logging, alerting, monitoring, or detection content for identifying suspicious or policy-violating behavior in applications or workflows.
Understanding of access control, identity boundaries, secrets handling, secure integration design, and environment-based deployment controls.
Ability to work with engineering teams to translate security concerns into implementable guardrails, validations, and release controls.
Strong written documentation and communication skills, especially for controls, findings, remediation evidence, and technical guidance.
Experience operating within enterprise governance, security, and release-management practices where evidence-based deployment readiness matters.

Nice To Haves

Experience with AI or agentic security controls, prompt and output protection strategies, or security validation of LLM-enabled features.
Experience with Microsoft, Azure, Copilot / Copilot Studio, or AI-enabled enterprise workflow platforms.
Experience with adversarial testing, red teaming support, detection engineering, or misuse-case validation for AI-enabled systems.
Experience in financial services, cybersecurity, regulated enterprise environments, or platforms with high audit and control requirements.
Familiarity with secure tool-calling patterns, API protections, model or prompt change validation, and runtime traceability for AI systems.
Working knowledge of cloud-native security patterns, telemetry analysis, and deployment gating for modern engineering teams.

Responsibilities

Implement and maintain security controls for AI-enabled applications, agents, prompts, and workflow automations, including input validation, output filtering, access controls, and governed tool-use restrictions.
Support development and operationalization of AI guardrails that reduce risk from prompt injection, unsafe responses, unauthorized tool execution, data leakage, and insecure workflow behavior.
Assist with adversarial and misuse-oriented testing activities by preparing scenarios, validating defenses, capturing findings, and supporting remediation with engineering teams.
Build and maintain monitoring and alerting logic for AI or agentic systems, including suspicious prompt patterns, abnormal workflow behavior, anomalous tool invocation, or policy-violating output patterns.
Contribute to runtime safety patterns such as content filters, role/permission boundaries, logging, evidence capture, and traceability controls required for secure deployment and audit readiness.
Partner with product and engineering teams to embed AI security expectations into design reviews, acceptance criteria, test plans, and deployment gates from the beginning of the delivery cycle.
Validate that AI-enabled solutions meet required security and governance standards before release, including deployment gate criteria, control validation, and documentation completeness.
Support incident investigation and remediation for AI-related control failures, suspicious behavior, or production observations that indicate safety or security degradation.
Help maintain control documentation, implementation notes, runbooks, remediation evidence, and operating procedures for AI security engineering activities.
Collaborate with AI quality, QA, platform, data, and security teams to ensure safe, resilient, and observable AI solutions across development and production environments.
Continuously improve automation, guardrail logic, validation workflows, and monitoring content as AI capabilities, workflows, and attack patterns evolve.