Back to jobs
Job Description
- Monitor Google Cloud AI products for signs of abuse, including prompt injection, jailbreaking, data poisoning, distillation, and generation of policy-violating content.
- Perform in-depth analysis of risks associated with both generative and agentic AI. Measure these risks using benchmarking, evaluations, red teaming, and scaled usage monitoring.
- Develop, tune, and deploy rules, heuristics, and rate limits to proactively block abusive actors and mitigate automated attacks.
- Effectively collaborate with engineering, product, and legal teams to ensure that the risks of AI are understood and robust solutions are adopted.
- Educate cross-functional teams about Gen AI safety risks and advocate for secure design principles. Promote a culture of safety and user trust throughout the product development process.