AI Penetrates and Blocks... Cyber Vulnerabilities to Double This Year

The rapid development of AI has significantly accelerated the discovery of software security vulnerabilities, leading to a projected doubling of publicly disclosed cyber vulnerabilities this year compared to last year. As AI enhances both offensive and defensive capabilities, the cybersecurity landscape is entering a new phase. According to the U.S. National Vulnerability Database (NVD), 45,207 software security vulnerabilities were reported between January and recently, nearly matching the total number recorded last year. If current trends continue, this year's vulnerability count could reach twice last year's record high. While increased vulnerability discovery indicates higher attack risks, it also reflects security firms and software companies leveraging AI to detect issues more rapidly. Major IT companies have also seen a surge in vulnerability remediation efforts, with Oracle releasing the most security patches in its 49-year history in July's update, Microsoft hitting a monthly record with 642 disclosed vulnerabilities, and Google fixing 433 vulnerabilities in its Chrome browser update—over 30 times higher than the 11 identified a year earlier. Industry experts highlight AI as the key driver behind these changes, with Google noting AI-powered systems identified 401 out of 433 vulnerabilities in Chrome. However, analysts note that the rapid increase in vulnerability discoveries does not necessarily mean cyber attack incidents are rising at the same pace.