Protecting people from harmful manipulation

This article provides a technical analysis of protecting individuals from harmful manipulation by AI systems, referencing a DeepMind blog post. It details a threat model, identifying potential actors and attack vectors such as data manipulation and deepfakes.