Google’s threat report puts the focus on faster AI-assisted attacks

Announcement

8 Sept 2026

Coverage

Global

Primary source

Google Threat Intelligence Group ↗

What changed?

The security story around AI is no longer just about convincing a chatbot to produce a bad answer. In its September 8 report, Google Threat Intelligence Group describes adversaries using agent-based workflows to automate parts of their operations.

Researchers say one observed campaign progressed from a compromised cloud resource to an agent-enabled credential-harvesting operation in under six hours. The report also discusses attempts to manipulate AI coding assistants and security scanners through compromised open-source software.

What should you keep in mind?

These are observations from Google’s investigations, not a measurement of every organisation’s exposure. Still, they show why the speed of detection and response deserves attention alongside prevention.