Vulnerability Discovery and Exploitation Trends in the AI Era (cloud.google.com)

The Google Threat Intelligence Group (GTIG) recently published findings indicating a significant shift in vulnerability discovery and exploitation trends attributed to artificial intelligence (AI). From January to August 2026, the monthly disclosure of vulnerabilities doubled, rising from 5,045 to 10,740, while exploitation rates increased nearly as steeply, averaging 18 vulnerabilities exploited per month compared to 10.5 in 2025. Notably, AI-driven discovery has led to a decrease in low-risk vulnerabilities and an increase in moderate and high-risk ones, particularly those capable of causing remote code execution (RCE). This surge in vulnerabilities poses substantial threats to cybersecurity, with GTIG anticipating continuous growth in both disclosures and exploitation rates. They emphasize the need for organizations to adopt a more strategic approach to vulnerability management, shifting from mass patching to threat-intelligence-driven triage. This shift involves employing targeted edge defenses and automated remediation to effectively manage the heightened risk associated with the rapidly evolving threat landscape—an evolution increasingly influenced by AI's capabilities in identifying and weaponizing vulnerabilities.