OpenAI just snagged an Anthropic safety researcher for its high-profile head of preparedness role

OpenAI has strengthened its safety division by appointing Dylan Scand, a notable AI safety expert previously with Anthropic, as its new head of preparedness. This pivotal role comes with a lucrative compensation package that could reach up to $555,000, along with equity options. The announcement, made by CEO Sam Altman on social media, highlights the urgency of addressing growing concerns around AI safety. Altman expressed his enthusiasm for Scand's arrival, stating, "Things are about to move quite fast and we will be working with extremely powerful models soon." He emphasized Scand's qualifications, declaring him the best candidate he has encountered for this critical position. In reaction to his new role, Scand expressed gratitude for his experiences at Anthropic and recognized the dual-edged nature of AI advancements, mentioning the significant benefits alongside the potential for severe risks. The position, described by Altman as "stressful," requires a leader capable of navigating complex technical challenges and making urgent decisions. OpenAI's job listing specified that candidates must possess extensive knowledge in machine learning, AI safety, and risk management, as well as the ability to unify diverse stakeholders on safety-related decisions. The move comes amidst growing scrutiny of OpenAI's safety practices, with several former employees, including the previous head of safety, having departed the company in recent years. Additionally, OpenAI is dealing with legal challenges from users who claim its tools have led to harmful behaviors. In a recent report, the company revealed that about 560,000 ChatGPT users each week exhibit signs of potential mental health issues, prompting them to consult mental health professionals to improve how the chatbot addresses these concerns.