Recent reports indicate that OpenAI has dissolved its preparedness team, a group that was responsible for evaluating potential risks associated with its models and finding ways to address those risks. This team had specifically focused on concerns like the models potentially acting unpredictably, such as compromising other companies’ security. Responsibilities have now been redistributed into existing teams, concentrating on specific areas such as biotechnology and cybersecurity.
This development comes amidst significant changes within the organization, which appears to be preparing for a major initial public offering (IPO). Over the past few years, OpenAI has been moving away from its previous comprehensive approach, dismantling its teams focused on artificial general intelligence (AGI) and superalignment. Notable figures including Ethics Lead Chloé Bakalar, Chief Futurist Josh Achiam, and Safety Head Johannes Heidecke have all departed from the company recently. This has raised concerns among critics, such as Jan Leike, who resigned in 2024, suggesting that OpenAI is prioritizing the launch of appealing products over ensuring safety measures.
