Applied AI Security Engineer: Frontier Threat Detection

swift; Greater London; Vor Ort; GBP 90.000 - 150.000; Vollzeit

Zusammenfassung

Swift, the global leader in secure financial messaging, seeks a frontier AI security researcher to run AI/LLM–driven vulnerability discovery against SWIFT codebases and integrate AI-assisted detection into development workflows.

Required are hands-on AI/LLM security experience, strong coding skills across languages, solid knowledge of common vulnerabilities and tooling, and a proven ability to build pipelines and document decisions for peer review.

Qualifikationen

• Hands-on experience applying AI/LLM models to security testing.
• Ability to read and reason about large codebases across languages.
• Understanding of vulnerability classes and analysis tools.
• Experience building tools and pipelines from scratch.

Aufgaben

• Run frontier AI models against SWIFT's codebases to find vulnerabilities.
• Develop the methodology through experimentation.
• Build an AI-assisted detection pipeline, dedupe findings, route issues to remediation.
• Run evaluations of frontier models on real detection tasks and decide which to rely on.
• Integrate AI-assisted detection into SWIFT software development with existing tooling.
• Identify vulnerability classes unseen by current tooling and build detection approaches.
• Work within governed access and tooling controls.
• Document methodology for peer scrutiny.

Kenntnisse

AI/LLM security

Code reading across languages

Application security fundamentals

Scripting & automation

Technical evaluation design

Ambiguity tolerance

Technical communication

Ausbildung

Bachelor's degree in CS/IS or related

Masters degree an asset

Jobbeschreibung

Swift, the global leader in secure financial messaging, seeks a frontier AI security researcher to run AI/LLM–driven vulnerability discovery against SWIFT codebases and integrate AI-assisted detection into development workflows.

Required are hands-on AI/LLM security experience, strong coding skills across languages, solid knowledge of common vulnerabilities and tooling, and a proven ability to build pipelines and document decisions for peer review.