Attackers Are Handing the Work to AI Agents, and Defenders Have Less Time to Respond

Google Threat Intelligence Group (GTIG) says that since its May 2026 report, forward-leaning attackers have moved from simple prompting to agentic AI workflows and AI-enabled automation. Because humans are less involved in each step, the window for defenders to respond is shrinking. In Q2 2026, GTIG saw attackers compromise a cloud resource, then plan, build and run an agent-enabled mass credential harvesting campaign in under six hours.

The software supply chain is also under pressure. GTIG tracked a group called UNC6780 using multiple tactics to trick AI coding assistants and large language model (LLM) security scanners as part of its open source supply chain compromises. Attackers are actively targeting developers, AI coding assistants and the scanning tools meant to protect them.

AI assets themselves are now valuable targets. GTIG observed adversaries going after proprietary AI models, code, prompts and research in sectors including healthcare, government and media. Others stole API credentials or took over victim cloud environments to run unauthorised AI workloads. State-sponsored and cyber crime groups also continue to use AI across the attack lifecycle, from reconnaissance and social engineering lures to malware obfuscation and troubleshooting after a break-in.