Karen Hao

Karen Hao is a journalist whose investigative reporting has shaped public understanding of the operational, environmental, and human costs of large-scale artificial intelligence. She is a contributing writer at The Atlantic, where her reporting focuses on the AI industry, and was previously the senior editor for AI at MIT Technology Review, where she founded the publication's AI coverage and built one of the most respected reporting beats in the field.

Hao earned her bachelor's degree in mechanical engineering from the Massachusetts Institute of Technology, training that has informed her ability to report on AI as a technical system rather than as a marketing narrative. Her engineering background allows her to interrogate the actual claims AI companies make about their systems and to cross-reference those claims against the engineering reality of how the systems are built, trained, and deployed.

Her 2019 MIT Technology Review article "Training a Single AI Model Can Emit as Much Carbon as Five Cars in Their Lifetimes" was among the first widely cited pieces of journalism quantifying the environmental cost of large-scale AI training, drawing on research by Emma Strubell and colleagues at the University of Massachusetts Amherst. The article is now a standard reference in discussions of AI and climate. Her subsequent reporting has investigated the labor conditions of data labelers in Kenya, Venezuela, and the Philippines who annotate training data for major AI systems; the operational practices and corporate culture of OpenAI, Anthropic, and other major AI labs; and the broader political economy of AI infrastructure.

Her 2025 book Empire of AI: Inside the Reckless Race for Total Domination is a longform investigative account of the AI industry, drawing on years of reporting and hundreds of interviews. The book examines how a small number of companies built infrastructure of unprecedented scale and what the human, environmental, and geopolitical costs of that buildout look like in practice.

Her reporting has been recognized with multiple journalism awards and is widely cited in academic, policy, and industry analyses of AI. She has been a guest lecturer at MIT, Harvard, and Stanford, and her work is frequently used as primary source material in AI policy and ethics courses.

Karen Hao's work grounds Discovery Before Disruption. The treatise cites her MIT Technology Review article on the carbon cost of training large AI models as primary evidence that AI deployment at scale has real-world consequences that must be understood before committing to a direction.

Her broader reporting extends that argument across multiple dimensions. AI deployment at scale produces measurable costs in energy consumption, water use, labor exploitation, and geopolitical leverage, costs that are typically invisible at the moment of deployment but real, durable, and globally distributed. The principle's directive to "remove only what you understand; build to discover the rest" has its journalistic case in Hao's body of investigative work, which makes the hidden infrastructure of AI legible to policymakers, researchers, and the broader public.