SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 63,000 - 90,000 / annual
Alice is a trust, safety, and security company built for the AI era. They are seeking a GenAI Analyst to join their team and play a vital role in securing generative AI tools by analyzing content infringements and identifying model vulnerabilities.
In this role, you will work at the intersection of AI safety and adversarial testing. Your primary responsibilities include developing adversarial and risky prompt strategies to expose weaknesses in various AI models, including Large Language Models (LLMs), Text-to-Image, Text-to-Video, and AI Agents. You'll collaborate with experts across diverse domains such as Hate Speech, Misinformation, Intellectual Property, and Copyright to understand abuse vectors and safety risks.
Key responsibilities include:
- Developing and executing adversarial prompt strategies across multiple abuse categories to identify model vulnerabilities
- Managing end-to-end projects from planning and oversight through quality assurance and delivery
- Handling extensive multilingual datasets with precision and meticulous attention to detail
- Investigating new tactics for circumventing foundational model safety measures
- Working cross-functionally with engineering, product, and policy teams to develop forward-thinking strategies
- Promoting knowledge sharing and continuous learning within the team
You must have a background in AI Safety, Responsible AI, or Trust and Safety. Familiarity with recent generative AI models and agents is essential, though direct technical experience is not required. Near-native English proficiency is mandatory, along with strong organizational skills and the ability to manage multiple concurrent tasks.
Desirable qualifications include experience with various model types (Text-to-Text, Text-to-Image) and prior OSINT (Open Source Intelligence) experience. A self-starter mentality and ability to thrive in a fast-moving environment are valued.
