Google’s threat intel chief explains why AI is now both the weapon and the target

Fast Company Tech

by Victor Dey

February 19, 2026

AI-Generated Deep Dive Summary

Generative AI has become a cornerstone of modern enterprise infrastructure, deeply integrated into software, cloud platforms, and internal processes. However, this shift has introduced new vulnerabilities, making AI systems themselves high-value targets for cyberattacks. According to Google Cloud’s AI Threat Tracker report, there is a growing trend of "model extraction" or "distillation" attacks, where attackers prompt generative AI models with carefully designed questions to reverse-engineer their capabilities. These attacks often exploit legitimate access points, making them harder to detect and shifting cybersecurity focus toward protecting intellectual property rather than just network perimeters. State-sponsored and financially motivated actors from nations like China, Iran, North Korea, and Russia are increasingly leveraging AI across the attack cycle. They use generative models to improve malware, craft phishing messages, mimic internal communications, and even assist with vulnerability discovery. These advancements allow attackers to operate at machine speed, overwhelming traditional human-driven defenses. For instance, AI can rapidly analyze data for targeting research and social engineering campaigns, significantly reducing the time and effort required for attacks. The integration of AI into cybersecurity is becoming a competitive arms race. While defenders race to deploy AI-based solutions to identify and respond to threats in real time, attackers are using AI to gain scale, speed, and sophistication. The potential for truly autonomous adversaries operating at scale is still emerging, but early signals suggest that threat velocity is accelerating. As AI becomes both a weapon and a target, the stakes are higher than ever, with foundation models representing billions in enterprise value. For those interested in design, this shift highlights the need to rethink cybersecurity strategies. AI’s role as a strategic asset requires new approaches to protect proprietary logic and intellectual property. Designers and developers must consider how AI systems can be made more resilient against reverse-engineering while maintaining their productivity and growth potential. The evolving landscape of machine-versus-machine threats underscores the importance of proactive measures to secure AI-driven infrastructure, ensuring that it remains a tool for innovation rather than exploitation.

Verticals

designtech

Originally published on Fast Company Tech on 2/19/2026