External Safety Testing Manager, Responsible Development and Innovation, DeepMind

Google•New York, NY

5h•$156,000 - $227,000

About The Position

As an External Safety Testing Manager on the Responsible Development and Innovation (ReDI) team, you will help deliver and scale external safety partnerships and evaluations for DeepMind’s models. You will work across Google and DeepMind with Product Management, Research, Legal, Engineering, Public Policy, Trust and Safety, and Frontier Safety and Governance teams. External safety evaluations are described as an important part of DeepMind’s safety practices and responsibilities.

Requirements

At least 8 years of experience working in a start-up, technology company, or consulting organization.
Experience working with AI or in an AI lab/start-up environment.

Nice To Haves

Experience designing and implementing testing and evaluations of cutting-edge AI systems.
Excellent communication skills and ability to work in cross-functional teams, foster collaboration, and influence outcomes.
Project management skills to optimize existing processes and create new processes.
Subject-matter expertise in generative AI safety considerations, including frontier safety risks such as chemical, biological, radiological, nuclear, cyber, harmful manipulation, and alignment risks.

Responsibilities

Lead frontier safety post-deployment testing strategy and execution.
Optimize and scale the post-deployment program to address novel risk and capability areas and changing external requirements, including identifying new partners with relevant skillsets.
Lead engagement with external testing partners and internal teams to execute the end-to-end testing process.
Engage internal subject-matter experts to assess the severity of findings and communicate them to leadership, domain experts, governance, and modeling teams to inform internal risk assessments and mitigations.
Work with Legal and Partnerships on contracting for new external testing partners.