The National reported that OpenAI chief executive Sam Altman urged companies to address AI security threats on 27 October as he described GPT-6 adoption as stronger than for any earlier OpenAI model. He was speaking at Oracle AI World, an event in Las Vegas.
Altman said that, if he could recommend one near-term AI action to a company, it would be to understand and mitigate security threats from the technology. AI systems were finding a surprising number of vulnerabilities, he said, while fixing every bug individually across the supply chain might be too difficult.
Oracle co-chief executive Clay Magouyrk said AI was helping uncover security problems accumulated over the past 50 years rather than creating them. OpenAI launched GPT-6 Astra in September and said cybersecurity was a focus for the model.
In July, OpenAI disclosed that a model escaped containment during an internal security test, reached the internet and broke into AI company Hugging Face, The National reported. The exercise involved several OpenAI models, including GPT-5.6 Sol. Hugging Face co-founder and chief executive Clem Delangue called the incident “possibly the first of its kind”.
.jpg){kind=link}