Datadog is looking to solve the problem of security in Generative AI technologies by researching and discovering vulnerabilities, attack vectors, and defensive strategies.
Requirements
- Deep knowledge of common Generative AI technologies and frameworks (MCP, A2A, LangGraph, CrewAI, etc.).
- Deep understanding of vulnerabilities specific to Generative AI, such as prompt injection, model poisoning, etc.
- Knowledgeable with the OWASP Top 10 for LLMs, and/or how common web security concepts (access controls, input handling, API security) apply to LLM models.
- Experience working in offensive security roles (penetration testing, red teaming) or vulnerability research with a focus on cloud or SaaS production environments/technologies.
- Comfortable with taking ambiguous requirements and creating a research plan, with measurable and tangible outcomes, as well as getting buy-in on the plan from stakeholders autonomously.
- Comfortable presenting and documenting your research findings with others: either internally at your current role, or publicly in conference talks and blog posts.
- Ability to write software to solve problems with and without AI tooling, and build systems for research purposes (eg. Go, Python, Rust, etc.).
Responsibilities
- Drive strategic research initiatives on Generative AI security by proposing, validating, and executing innovative projects.
- Conduct hands-on research to discover and demonstrate vulnerabilities, attack vectors, adversarial methods, and misconfigurations in Generative AI and large language model (LLM) technologies.
- Create proof-of-concept attacks, simulations, and demonstrations to illustrate vulnerabilities and defensive strategies clearly.
- Serve as a subject matter expert in Generative AI and collaborate with product management, engineering, and detection engineering teams to translate research findings into actionable product improvements.
- Author and present impactful blog posts, webinars, and conference presentations to educate the broader security and AI community.
- Engage closely with cloud providers, AI vendors, and open-source communities to responsibly disclose and remediate identified security issues.
- Track, research and experiment with the latest tactics, techniques & procedures for attacking and defending Generative AI infrastructure.
Other
- Bachelor's, Master's, or Ph.D. degree in a relevant field.
- Ability to work in a hybrid workplace and collaborate with team members.
- Strong communication and presentation skills, both verbal and written.
- Ability to work in an OKR driven environment.
- Experience working in a Security Research organization.
