Red Teaming artificial intelligence for social good - The PLAYBOOK

As Generative Artificial Intelligence becomes an integral part of our digital landscape and daily life, understanding its risks and participating in solutions is crucial to ensuring that it works for the overall social good. Building on a Red Teaming event co-hosted by UNESCO and Humane Intelligence, this PLAYBOOK guides the reader through the Red Teaming process, where participants test Gen AI models for flaws and vulnerabilities that may uncover harmful behaviour. It is a practical guide designed to empower organizations, researchers, decisionmakers and communities to systematically test Generative AI models. To illustrate potential harms, the PLAYBOOK highlights how these models can inadvertently reinforce stereotypes and biases—and, in some cases, be intentionally leveraged to enable technology-facilitated gender-based violence (TFGBV) at an unprecedented speed and scale.