Timnit Gebru

Timnit Gebru is an AI researcher, computer scientist, and one of the most influential voices in the field of AI ethics, bias, and accountability. She is the founder and Executive Director of the Distributed AI Research Institute (DAIR), an independent research organization she founded in 2021 to conduct AI research that centers the perspectives and needs of communities most affected by AI systems, rather than the priorities of large technology corporations.

Gebru earned her PhD in computer vision from Stanford University and subsequently joined Microsoft Research as a researcher before moving to Google Brain as co-lead of its Ethical AI team. Her departure from Google in December 2020 became one of the most prominent controversies in the history of AI ethics: she was dismissed, in circumstances Google described as a resignation and she described as a firing, after Google management objected to the publication of a research paper on the environmental and bias risks of large language models. The event drew significant attention to the governance of AI ethics research inside large technology companies and led to wide coverage and a broader conversation about the independence of AI safety and ethics researchers.

Before founding DAIR, Gebru co-authored several highly cited papers establishing methodological frameworks for AI accountability, including "Datasheets for Datasets," which introduced a standardized documentation framework for training datasets, and "Model Cards for Model Reporting," which proposed a similar approach for AI model documentation.

Her work on facial recognition bias, particularly research showing that commercial facial recognition systems from major vendors performed significantly worse on darker-skinned and female faces, contributed directly to policy discussions about the regulation of biometric AI.

Gebru is also a co-founder of Black in AI, an organization established to increase the representation of Black researchers and practitioners in artificial intelligence and to support their work through community, mentorship, and conference programming. The organization has been instrumental in shifting the demographic and intellectual composition of the AI research community at major venues. Through DAIR and her broader public work, Gebru has consistently argued that AI accountability requires institutional independence: research conducted inside the major AI labs and platform companies cannot reliably hold those organizations accountable, because the incentives, reporting lines, and legal constraints inside those companies systematically prevent unflattering findings from being published. DAIR is her structural answer to that problem, and the AI First Principles' insistence on principles that survive corporate convenience reflects the same diagnosis. She has spoken before legislators, in international policy forums, and at academic and industry conferences on the regulatory and institutional design questions that determine whether AI can be developed in the public interest.

Timnit Gebru's "Datasheets for Datasets" paper is cited directly in the treatise as the solution framework for AI Inherits Messiness. The treatise draws on her work to make the principle's central argument concrete: "the solution is not sanitizing training data but understanding and documenting what biases it contains, then designing systems with guardrails that prevent those biases from causing harm."

Her contribution reframes the problem from a technical defect to a documentation and accountability practice. Every dataset carries the context of its creation — who made it, under what circumstances, and what assumptions were considered normal at the time. Making those conditions visible is not optional; it is the prerequisite for building AI systems that do not systematically amplify the prejudices of their training data. Gebru's work provides the methodology for doing so.