A shocking discovery has been made about the potential vulnerabilities of OpenAI's cutting-edge AI agent, Codex. Researchers have found that the model can be used to launch devastating denial-of-service (DOS) attacks on web servers, leaving users worried about the security of their online services.
The vulnerability lies in a decade-old weakness in the HTTP/2 protocol, which was exploited by hackers to crash web servers in seconds. According to experts, this type of attack can cause significant downtime and financial losses for businesses and individuals alike.
OpenAI has confirmed that Codex is capable of launching DOS attacks, but insists that it did not knowingly create the vulnerability. The company says it is working closely with researchers and security experts to address the issue and prevent similar attacks in the future.
The implications of this discovery are far-reaching, with many businesses and organisations depending on web services for their operations. The UK's Information Commissioner's Office (ICO) has issued guidance on protecting against DOS attacks, advising organisations to implement robust cybersecurity measures to mitigate potential risks.