Join the pioneering Content Adversarial Red Team (CART) within Google's Trust and Safety division. This innovative team focuses on unstructured adversarial testing of cutting-edge Generative AI products. Your role will be crucial in uncovering novel content risks that might be missed by traditional evaluation methods. By collaborating closely with product, policy, and enforcement teams, you'll directly contribute to building the most secure and trustworthy AI experiences for Google users worldwide.

As a vital member of CART, you will spearhead the identification and mitigation of emerging content safety challenges in Google's Generative AI offerings. You'll independently manage red teaming projects from conception to completion, proactively discovering unknown Generative AI issues, novel threats, and vulnerabilities that standard testing protocols may overlook. Your advanced analytical abilities, precise testing execution, and talent for devising creative attack strategies will be paramount in ensuring Google's AI products meet the highest standards of safety and reliability before public release.

Lead end-to-end red teaming initiatives focused on uncovering novel content safety risks within Google's core Generative AI products. Develop and implement sophisticated red teaming strategies, alongside designing targeted, comprehensive testing for specific product scopes. Collaborate with product managers, engineers, researchers, and Trust & Safety stakeholders to deeply understand product functionalities, potential vulnerabilities, and jointly develop effective mitigation strategies. Analyze raw attack outcomes and perform detailed qualitative assessments of complex edge cases. Translate raw vulnerability data into clear, actionable insights and identify critical risks. Contribute significantly to the advancement of AI safety methodologies and operational excellence through the development and execution of designated CART programs. Critically evaluate existing project methodologies, identify procedural gaps, and refine testing playbooks and Standard Operating Procedures (SOPs) to amplify CART's impact.

A Bachelor's degree or equivalent practical experience is required for this role.

You should possess a minimum of 4 years of experience in adversarial testing, red teaming, or fields related to GenAI/AI safety, ethics, or responsibility.

Preferred qualifications include a Master's degree in a relevant discipline such as Computer Science, Information Security, or Artificial Intelligence.

Experience with Google's products and services, particularly its GenAI offerings, is highly beneficial.

Familiarity with working with large datasets, data analysis tools, and an understanding of various AI model architectures and their potential vulnerabilities are advantageous.

Demonstrated ability to work effectively both independently and as part of a collaborative team is essential.

Exceptional problem-solving and critical thinking skills, coupled with meticulous attention to detail, are necessary to thrive in this dynamic and evolving environment.