Structural turbulence reappears as they rush for IPO. OpenAI is reported to have disbanded its security preparation team. According to the Financial Times, artificial intelligence giant OpenAI has officially disbanded its security preparation team (preparedness team) at the end of last month.

Structural shock reappears as the company rushes for an IPO. OpenAI is reported to have disbanded its security preparation team. According to the Financial Times, artificial intelligence giant OpenAI has officially disbanded its security preparation team (preparedness team) at the end of last month. The team's previous main responsibility was to comprehensively assess whether artificial intelligence models pose serious security risks, such as the possibility of the model getting out of control and launching hacking attacks on other companies, and to develop mitigation solutions for these potential risks. According to reports, after the team was disbanded, its original responsibilities have been further split into more specific areas such as biotechnology and cybersecurity, and have been reorganized and merged into other existing business teams of the company. This structural adjustment is just the latest in a series of recent internal turmoils at OpenAI, which is currently preparing for an expected large-scale initial public offering (IPO). In the past few years, OpenAI has gradually broken its original research-led model that was more inclined to extensive exploration, and successively disbanded the AGI readiness team (AGI readiness) and the super alignment team (superalignment). Along with frequent team reorganizations, the company's top management has also experienced significant personnel changes. Recently, a number of key figures have resigned, including head of ethics Chloé Bakalar, chief futurist Josh Achiam, and head of security Johannes Heidecke. This series of actions has triggered strong concerns among some industry insiders and former employees. Jan Leike, who resigned from OpenAI earlier in 2024, bluntly criticized the company in an interview with the Financial Times, saying that the company is currently leaving crucial security issues behind in order to launch "shiny new products". In addition, the whereabouts of the original preparation team leader have also been clarified. Dylan Scandinaro, who was poached by OpenAI from competitor Anthropic in February this year, will completely change his focus after the team is disbanded. It is reported that his future work direction will focus on exploring and studying the far-reaching impact of "recursive self-improving" artificial intelligence. via cnBeta.COM - Chinese industry information station (author: source: cnBeta.COM)