OpenAI announced on 24 March 2026 that it is establishing a new entity called the OpenAI Foundation, marking a structural change to the organisation originally founded as a non-profit research laboratory in 2015 [source].

The company stated that the Foundation will hold oversight responsibilities and that OpenAI's existing for-profit subsidiary will continue commercial operations under a revised governance framework. OpenAI did not specify whether the Foundation will have binding authority over product decisions, safety protocols, or model deployment timelines.

The announcement follows years of public debate over OpenAI's 2019 transition to a "capped-profit" structure, which allowed external investment while maintaining stated commitments to beneficial AI development. Critics have questioned whether commercial incentives have influenced research priorities and safety practices.

OpenAI said the Foundation will be funded separately and will focus on "long-term AI safety and alignment research." The company did not disclose the Foundation's budget, governance composition, or whether it will publish independent audits of OpenAI's systems.

The move comes amid increased regulatory scrutiny of AI providers in multiple jurisdictions. The European Union's AI Act entered provisional application in February 2026, and the United States Federal Trade Commission has opened inquiries into several major AI companies' data practices and model behaviour.

OpenAI stated that further details about the Foundation's structure and leadership will be released "in the coming months." The company did not address whether the Foundation will have authority to halt or modify deployments of existing models, including GPT-4 and its successors, if safety concerns arise.