Google says attackers used AI agents to steal credentials in under six hours
Under six hours. That's how long attackers used AI agents to steal thousands of credentials — and Google just documented how.
Why it matters
Real-world agent exploitation is no longer theoretical: threat actors are assembling autonomous AI frameworks for coordinated attacks at scale. This is the first major documented case of multi-agent compromise, signaling that agent security is now a critical operational concern for every enterprise.
The key facts
7 to knowMulti-agent AI framework used to compromise thousands of credentials
Attack completed in under 6 hours
Campaign traced to financially motivated threat actor
Initial compromise via cloud infrastructure breach
Autonomous framework assembled post-breach for lateral movement/data theft
Google Threat Intelligence Group / Mandiant investigation
Published September 8, 2026
Go to the source
SiliconAnglesiliconangle.com
Publisher excerpt: Threat actors used a multi-agent artificial intelligence framework to compromise thousands of credentials in under six hours, Google LLC’s Google Threat Intelligence Group said in a report released today. Mandiant investigators traced the campaign to a suspected financially motivated actor that…