OpenAI Begins Training GPT-5 Successor, Forms New Board Committee for Safety Oversight

OpenAI, the artificial intelligence research lab at the forefront of the generative AI boom, has officially announced that it has begun training its next major AI model, widely anticipated to be the successor to the powerful GPT-4. In a concurrent move aimed at addressing mounting concerns over AI safety and governance, the company also revealed the formation of a new Safety and Security Committee led by members of its board of directors.

The new committee is tasked with making critical recommendations to the full board regarding the safety and security of all OpenAI projects. It will be responsible for evaluating and evolving the company’s processes and safeguards as it develops increasingly capable systems. The committee is chaired by Bret Taylor, Chairman of the OpenAI board, and includes CEO Sam Altman, as well as board members Adam D’Angelo, Nicole Seligman, and recently appointed former U.S. National Security Agency (NSA) Director, Gen. Paul M. Nakasone. OpenAI has committed the committee to delivering its first set of recommendations within the next 90 days.

This announcement comes at a pivotal time for the company. It follows the recent departure of key safety-focused leaders, including co-founder Ilya Sutskever and Jan Leike, who publicly warned that safety culture and processes at OpenAI had “taken a backseat to shiny products.” The formation of the new board-led committee appears to be a direct response to these criticisms and a broader effort to reassure the public and policymakers that safety remains a core priority.

While details about the next-generation model remain scarce, the commencement of its training signals OpenAI’s intent to maintain its competitive edge against rivals like Google, Anthropic, and Meta. The dual announcements highlight the central tension in advanced AI development: the relentless push for more powerful models balanced against the profound responsibility of ensuring they are developed and deployed safely. The work of the new committee will be closely watched as a bellwether for how the industry’s leader intends to navigate the path toward artificial general intelligence (AGI).