The quality of personas hinges on robust user data, and poorly researched or assumption-based personas can exacerbate problems rather than solve them. Generative AI, such as ChatGPT and Google Bar, holds promise for enhancing persona development by incorporating user data from a wider range of sources (e.g., social media), but the extent to which it can be relied upon as an accurate source of data is unknown because generative AI lacks the capacity to understand the context and provide real-world examples of human experiences. Our presentation explores the potential of generative AI in persona development, comparing personas created with AI to researcher-generated personas for a water supply forecasting system. We assess AI’s impact on persona authenticity and effectiveness and offer empirical evidence on the benefits and limitations of using AI to augment persona development, shedding light on its role in improving user-centered design practices.