Google Introduces Agentic AI to Combat Cybersecurity Threats

Google’s M-Trends 2025 summary reports 75 zero-day exploits tracked in 2024 (a slight year-over-year dip but within an overall rising trend), with exploits as the leading initial access vector and stolen credentials increasing; enterprise/security/networking products comprised 44% of exploited zero-days and nation-state and commercial espionage actors were responsible for over half of attributed cases. In response, Google unveiled Unified Security and agentic AI capabilities (Curated Detections, Applied Threat Intelligence Rule Packs, alert triage and malware analysis agents) to help defenders detect and automate responses to threats including infostealers and cloud compromises.