Remote (Global)

LILT is hiring an AI Red Team Engineer - Traditional Chinese

About the Role

The AI Red Team Engineer will proactively test artificial intelligence systems to uncover weaknesses, especially in language models processing Traditional Chinese. This role involves designing attack simulations, evaluating model responses, and working with development teams to strengthen system defenses.

Responsibilities

  • Simulate adversarial behaviors to expose flaws in AI models
  • Conduct penetration testing on language processing systems
  • Analyze model outputs for security and accuracy issues
  • Develop strategies to improve model resilience
  • Test AI behavior under edge-case scenarios
  • Document vulnerabilities and recommend fixes
  • Collaborate with engineering teams on mitigation
  • Evaluate model performance in multilingual contexts
  • Focus on Traditional Chinese language patterns and nuances
  • Assess risks related to misinformation and bias
  • Monitor emerging threats in AI applications
  • Improve detection of malicious inputs
  • Validate model responses against intended behavior
  • Support red teaming frameworks and tools
  • Report findings to technical leadership
  • Maintain up-to-date knowledge of AI security trends
  • Ensure compliance with ethical guidelines
  • Test for prompt injection and data leakage
  • Evaluate cross-lingual transfer of vulnerabilities
  • Assist in training defensive models
  • Work with natural language understanding systems
  • Identify weaknesses in model reasoning
  • Stress-test models under adversarial conditions
  • Contribute to secure AI development practices
  • Support continuous improvement of model safety

Compensation

Competitive salary and benefits package offered.

Work Arrangement

Remote position with flexible scheduling options.

Team

Collaborative team environment focused on AI safety and language technology development.

Language Focus

  • Primary focus on Traditional Chinese language inputs and outputs
  • Evaluate cultural and linguistic nuances in model responses
  • Test for regional variations in Chinese dialects and usage
  • Assess translation accuracy and context preservation
  • Monitor for inappropriate or harmful content in Chinese text

Security Testing Scope

  • Test for prompt injection and command override
  • Evaluate resistance to jailbreaking attempts
  • Assess data leakage through model outputs
  • Check for unintended model behaviors
  • Validate input sanitization mechanisms

Visa sponsorship available for qualified candidates.

Required Skills
PythonBashPowerShellDockerGhidraRed TeamingVulnerability AssessmentPenetration TestingReverse EngineeringThreat ModelingSecurity ToolingScriptingMalware AnalysisAdversarial AI
About company
LILT
LILT builds multilingual AI and human-verified services that make the world's information available to everyone, regardless of language. The company serves Enterprises, Governments, and AI Developers worldwide.
All jobs at LILT Visit website
Job Details
Category other
Posted 6 months ago