Sarah Barrington warns of AI-driven disinformation risks

Sarah Barrington highlights how artificial intelligence has supercharged the long-standing human tendency to alter content and rewrite history. While narrative manipulation has existed for centuries, the transition from machine learning to AI has increased the scale, speed, and ease with which perceptually realistic content can be created by average users.

Barrington identifies disinformation as a primary catastrophic harm, particularly in political contexts where the goal is to sow uncertainty and pollute the information ecosystem. This creates a secondary effect known as the ‘liar’s dividend,’ a state where the public questions everything to the point of being unable to distinguish between what is real and what is fake, potentially undermining legitimate news outlets.