AI Disaster Response Team Disbanded by OpenAI

OpenAI Disbands Preparedness Team Amidst IPO Preparations

A recent report by the Financial Times revealed that OpenAI, a leading artificial intelligence research company, has disbanded its preparedness team. This team was responsible for assessing potential risks posed by AI models and developing strategies to mitigate those risks, such as the threat of rogue behavior and hacking.

However, OpenAI spokesperson Kayla Wood refuted this claim in a statement to The Verge, stating that the Preparedness team remains intact. She emphasized the company’s strong research leadership in cybersecurity, biological and chemical fields, and AI self-improvement, all under the supervision of Saachi Jain, the head of safety.

Instead of disbanding the Preparedness team, OpenAI has reorganized responsibilities into specialized areas such as biosecurity and cybersecurity, which have been integrated into existing teams.

This restructuring is part of a series of changes within OpenAI as the company prepares for a major IPO. In recent years, OpenAI has shifted away from its reach-led model, dissolving teams focused on artificial general intelligence readiness and superalignment. Notable figures like ethics lead Chloé Bakalar, chief futurist Josh Achiam, and head of safety Johannes Heidecke have also departed, raising concerns among critics about the company’s safety priorities.

Former OpenAI employee Jan Leike, who resigned in 2024, criticized the company in the Financial Times, accusing it of prioritizing product development over safety considerations.

Update, August 18th: Additional response from OpenAI included.