World Prepares for New Catastrophe, All Urged to Be on Full Alert
OpenAI co-founder Greg Brockman has issued a warning about the growing cyber security threat as artificial intelligence (AI) continues to advance.
Brockman urged companies to immediately strengthen their security systems to counter cyberattacks supported by AI technology.
He said the hacking incident involving the AI developer platform Hugging Face marked a significant turning point in the cyber security landscape.
“I have spoken with many organisations over the past few weeks, and one theme is clear: they know they need to fundamentally upgrade their cyber security practices at an unprecedented pace,” Brockman wrote in his personal blog, as quoted by Business Insider on Tuesday (18/8/2026).
The warning came after OpenAI revealed in July that its AI agent had managed to escape its testing environment during an internal test. The system then successfully infiltrated Hugging Face, a platform used by developers to publish, share, and download AI models.
According to Brockman, the incident served as a clear example of how AI can be used to identify security vulnerabilities in systems more quickly.
He predicted that AI’s ability to detect weaknesses would continue to improve. However, the same technology could also help companies fix those flaws before hackers exploit them.
“AI will also make it far easier to patch security holes to prevent attacks,” Brockman wrote.
Brockman viewed the OpenAI-Hugging Face incident as a crucial moment for companies to strengthen their cyber defences. He even shared a list of ten steps that he believes companies on the defensive side of cyberattacks need to take immediately.
“Time is of the essence, and defenders need to execute the steps below at turbo speed,” he wrote.
According to Brockman, companies can no longer rely on conventional cyber security approaches as AI capabilities develop ever faster.
He argued that organisations must begin significantly automating their security programmes within the next few months to remain capable of facing the threat.
“In the coming months, every organisation will need to start significantly automating their security programmes to stay safe, and the security community must quickly step up to define the tools, practices, and playbooks that will increase defenders’ strength faster than attackers as AI continues to evolve,” he explained.
Source: jawawa.id (https://jawawa.id/newsitem/world-prepares-for-new-catastrophe-all-urged-to-be-on-full-alert-1787065196)
Originally published by CNBC Indonesia (https://www.cnbcindonesia.com/tech/20260818135646-37-760309/dunia-bersiap-hadapi-petaka-baru-semua-diminta-siaga-penuh), translated from Indonesian