Anthropic CEO warns: AI uncovers years-long hidden software vulnerabilities, triggering a cybersecurity crisis

Golden Finance reports that on May 6th, Anthropic CEO Dario Amodei warned that AI has uncovered a large number of long-hidden software vulnerabilities, and global tech companies, governments, and financial institutions are facing a critical moment. They must complete repairs within a limited time; otherwise, it could trigger widespread cybersecurity attacks and financial losses. Amodei pointed out that Anthropic’s latest AI model Mythos has discovered tens of thousands of software vulnerabilities during testing, some of which have existed for decades and involve critical core software systems. He stated that AI is significantly amplifying cybersecurity risks, with the number of vulnerabilities, intrusion incidents, and ransomware-related financial damages to schools, hospitals, and even banks potentially experiencing substantial growth. (Dongxin News)