Senior GenAI Safety Researcher
Alice is seeking an experienced and highly motivated Senior Generative AI Safety Researcher to join its U.S. team.
This role sits at the forefront of AI Safety and Trust & Safety research, focusing on identifying vulnerabilities, harmful behaviors, and content-related risks within advanced generative AI systems.
Rather than simply executing predefined tests, you will help design comprehensive evaluation methodologies and red-teaming frameworks capable of uncovering weaknesses in emerging AI technologies.
Your research will cover multiple AI modalities, including Large Language Models (LLMs), text-to-image systems, text-to-video models, multimodal systems, and autonomous AI agents.
You will also serve as a key subject-matter expert, supporting program leadership while collaborating with engineering, product, policy, research, and external stakeholder teams.
Key Responsibilities
AI Safety Research & Testing
- Design rigorous and scalable evaluation methodologies for testing foundation models, multimodal AI systems, and AI agents.
- Develop advanced red-teaming frameworks to identify weaknesses in model safeguards and safety mechanisms.
- Create sophisticated prompt-testing strategies covering different categories of AI risk.
- Evaluate potential vulnerabilities involving misinformation, hate speech, copyright and intellectual-property concerns, child safety, and other harmful-content categories.
- Research emerging approaches used to bypass or circumvent generative AI safety mechanisms.
- Investigate evolving jailbreak, prompt-injection, and adversarial prompting techniques.
- Translate research findings into practical recommendations that can improve model safety.
Research Strategy & Subject-Matter Expertise
- Serve as a trusted AI Safety and content-risk specialist for internal teams and program leadership.
- Provide technical and policy insights when defining project scope, assessing potential risks, and developing research strategies.
- Help establish consistent evaluation standards and methodologies across AI Safety projects.
- Document research findings and contribute to the continued development of Alice’s internal AI Safety knowledge base.
- Develop and maintain best practices, risk taxonomies, testing frameworks, and research documentation.
- Stay informed about emerging developments within Generative AI, model safety, adversarial testing, and Trust & Safety.
Team Development
- Provide guidance and mentorship to junior researchers and analysts.
- Review research and evaluation work to maintain high standards of accuracy and consistency.
- Help strengthen the team’s analytical capabilities and research practices.
- Encourage continuous learning and knowledge sharing across the organization.
Project & Operational Management
- Take ownership of research engagements throughout the complete project lifecycle.
- Manage projects from initial planning and methodology development through testing, quality assurance, analysis, and final delivery.
- Oversee complex datasets involving multiple languages and different categories of harmful or abusive content.
- Maintain high levels of precision, accuracy, and research quality.
- Coordinate effectively across multiple teams to ensure research objectives are achieved.
Cross-Functional Collaboration
- Partner with engineering, product, policy, research, and client-facing teams.
- Clearly communicate complex research findings to both technical and non-technical stakeholders.
- Help translate identified vulnerabilities into practical mitigation strategies.
- Support external stakeholders and clients by providing expert guidance on Generative AI safety risks.
Requirements
Candidates should have at least 5 years of professional experience in areas such as:
- AI Safety.
- Responsible AI.
- Trust & Safety.
- Generative AI research.
- AI evaluation.
- Closely related research disciplines.
The successful candidate should also demonstrate:
- Strong experience designing qualitative or quantitative evaluation methodologies for Generative AI systems.
- Deep understanding of AI-related content risks, including toxicity, misinformation, copyright concerns, harmful content, and safety-policy violations.
- Demonstrated ability to independently manage complex research projects from planning through final delivery.
- Excellent attention to detail and the ability to maintain research quality in rapidly changing environments.
- Strong familiarity with modern Generative AI architectures and technologies.
- Practical knowledge of prompt engineering and adversarial AI evaluation.
- Experience with AI red-teaming methodologies.
- Strong understanding of AI agents and emerging agentic systems.
- Excellent written and verbal communication skills.
- Ability to serve as a subject-matter expert for program leaders, internal teams, clients, and other stakeholders.
Preferred Qualifications
Additional experience that may strengthen your application includes:
- Published AI or technology research through academic publications, industry whitepapers, research organizations, or similar channels.
- Hands-on experience evaluating multimodal Generative AI systems, including text-to-image, text-to-video, and audio models.
- Experience mentoring or supervising junior researchers and analysts.
- Previous responsibility for reviewing or quality-assuring AI research and evaluation work.
- Strong familiarity with emerging AI Safety research, model vulnerabilities, and adversarial testing techniques.