AI Defense Engineer

WilmerHale; Boston (MA); On-site; USD 140,000 - 210,000; Full time

Job summary

WilmerHale seeks a Senior AI Defense Engineer to lead security direction for AI/ML systems in a global law firm setting. You will translate AI threats into defenses, develop guardrails, and drive policy enforcement, monitoring, and secure-by-default practices across engineering teams.

Responsibilities include threat modeling, risk assessments, and evaluating commercial AI security solutions to meet legal-sector requirements such as confidentiality and data residency.

Qualifications

• Experience designing and reviewing security controls for AI/ML systems and services.
• Ability to translate emerging AI threats into practical defenses, guardrails and policy layers.
• Proven ability to mentor teams and influence security direction in large, complex environments.

Responsibilities

• Lead threat modeling for AI/ML systems and identify AI-specific attack vectors.
• Evaluate and operationalize security controls, guardrails, and enforcement mechanisms for AI services.

Skills

Threat modeling

AI/ML security

Risk assessment

Security controls

Adversarial testing

Telemetry & monitoring

Job description

WilmerHale is a leading, full-service international law firm with 1,000 lawyers located throughout 12 offices in the United States and Europe. Our lawyers work at the intersection of government, technology and business, and we remain committed to our guiding principles of providing quality, excellent legal and client services; developing diversity among our lawyers and staff and cultivating an environment that promotes an ambitious spirit, collaboration and collegiality by drawing on the extraordinary talents and dynamic experience of our lawyers. Our goal is to reflect the diversity of our clients and the communities in which we practice.

** About the Role**

The Senior AI Defense Engineer is a technical leader responsible for securing AI in a global law firm environment. This role is responsible for setting technical direction, driving delivery, and mentoring colleagues to raise their awareness and capabilities. The role will translate emerging AI threats into practical defenses, guardrails, policy enforcement layers, monitoring and detections, adversarial test automation, and hardened environments that hold up under real attacker pressure.

The role will support a smart-integration, buy-before-build, security strategy. You will evaluate, select, and operationalize commercial Al security solutions that meet stringent legal-sector expectations, including matter confidentiality, ethical walls, client audit requirements, data residency constraints, and contractual information technology service obligations.

Success looks like: The role also enables progress by enhancing and performing commercial AI tool evaluations and approvals, assessing internally developed AI solutions, and responding to growing audit demands with credible evidence of AI cybersecurity protections. Additionally, success includes secure-by-default adopted by engineering teams; adversarial evaluation and assessments that reliably finds issues before production; telemetry and detections that catch abuse early; and an AI security roadmap that stays current with fast-moving technology shifts.

What You Will Be Doing

• Threat Modeling & Risk Assessment - Guide and conduct technical threat modeling for AI/ML systems (neural networks, expert systems, retrieval-augmented generation, classification models, etc.). Identify and document AI-specific threats with emphasis on how vendor controls (gateways, content filters, policy engines, etc.) mitigate prompt injection, data leakage, jailbreaks, and unsafe autonomy. Provide clear, prioritized mitigation guidance to colleagues via vendor configuration standards, reference patterns, and exception processes.
• AI Defense Engineering – Evaluate and operationalize security controls, guardrails, and enforcement mechanisms for AI services (e.g., input/output filters, policy enforcement layers, content safety checks, rate limiting, abuse detection). Enable detections and monitoring for AI-specific attack patterns using logs, telemetry, and model signals. Work with platform teams