AI Security Engineer: Guardrails & Identity Controls

Plaid; New York (NY); On-site; USD 189,000 - 243,000; Full time

Benefits offered by this job

Equity compensation

Commission opportunity

Comprehensive benefits (medical,Dental

Vision, 401(k)

Job summary

Plaid is seeking a security-focused engineer to be the first line of defense for AI security. You will review AI tech for risks, design guardrails, and implement identity controls to enable safe, aggressive AI use.

You’ll drive security reviews and craft scalable safeguards across the Plaid AI surface area. Ideal candidates have hands-on experience securing AI technologies, with at least one in the listed capabilities, and 3–6 years in security roles.

Qualifications

• Must-have: hands-on experience securing AI technologies, including at least one of the listed capabilities.
• Deploying guardrails or sandboxes,
• Managing AI identities,
• Detecting or mitigating prompt injection,
• Running AI security reviews,
• Implementing AI security tooling (built or bought)

Responsibilities

• Review AI systems and integrations for security risk before and after deployment.
• Design and deploy guardrails, sandboxes, and other controls that contain what AI agents can do.
• Build or implement systems to track and manage AI identities, so we know what access every AI has been granted and can balance risk with agility.
• Detect and mitigate prompt injection across Plaid's AI surface area.
• Run AI-focused security reviews and set the bar other teams build to.

Skills

AI security

Guardrails / Sandboxes

AI identities

Prompt injection

AI security reviews

Security tooling

3-6 years experience

Tools

AI security tooling

Job description

Plaid is seeking a security-focused engineer to be the first line of defense for AI security. You will review AI tech for risks, design guardrails, and implement identity controls to enable safe, aggressive AI use.

You’ll drive security reviews and craft scalable safeguards across the Plaid AI surface area. Ideal candidates have hands-on experience securing AI technologies, with at least one in the listed capabilities, and 3–6 years in security roles.