This critical review examines Karen Hao’s Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI as an institutional case study of mission-driven governance under commercial and capital pressure. Drawing on Hao’s extensive documentation of OpenAI’s governance structure, leadership conflicts, financial dependencies, outsourced data labour, and environmental resource consumption, the review explores the gap between formal authority and operational power within influential artificial intelligence organisations. It argues that the book is particularly effective in demonstrating how dependence on external capital can contribute to mission drift, weaken nonprofit oversight, and transfer labour and environmental risks to communities in the Global South. Although Hao’s imperial framework successfully identifies recurring patterns of extraction involving data, labour, land, water, and energy, it is less effective in explaining the institutional mechanisms through which such harms may be prevented or remedied. The review further considers the book’s relevance to universities and other organisations adopting generative AI systems, particularly in relation to procurement, vendor dependency, data governance, academic freedom, and accountability. It concludes that the book’s lasting contribution lies in its detailed documentation of how organisational missions, governance structures, and public-interest commitments can be reshaped by the resources required to pursue technological advancement.