OpenAI's Astra AI Can Independently Discover and Exploit Cyber Vulnerabilities

OpenAI has developed a new AI model called Astra that has reached the company's highest cybersecurity risk threshold, classified as 'Critical'. Astra is capable of identifying previously unknown security vulnerabilities and crafting methods to exploit them without human assistance. It is the first OpenAI model to achieve this level, meaning it can operate against hardened and well-protected systems. The development raises significant concerns about the potential misuse of advanced AI in cyberattacks.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.




Discussion (0)
Log in to join the discussion and vote.
Log in