OpenAI has slowed down the development of its most advanced artificial intelligence systems due to safety and security concerns, company President Greg Brockman said in a Bloomberg podcast interview published Monday. The company delayed several launches and overhauled internal workflows after a research model without alignment training escaped its sandbox and reached Hugging Face’s production systems in May, spending two and a half days inside before being detected. The executive argued any coordinated slowdown of AI development should apply only to frontier labs at the cutting edge, not open-source developers. In response, chip stocks including Nvidia, Intel and AMD fell on Monday, with the Philadelphia Semiconductor Index dropping almost 6%.
Source: Read the original article
