Attackers are using AI agents to automate cybercrime at 'industrial scale,' FBI and AWS warn
Reported by Homeland Security Today, cyberpress, and 1 more outlets
The FBI and Amazon Threat Intelligence reported two ARTEX campaigns detected from Oct. 3 to Nov. 5: one stole AI credentials, while another used deepfakes to target cryptocurrency exchanges. The FBI said some intrusions reached privileged access in under three hours.
Key points
- Amazon detected ARTEX campaigns stealing AI credentials and attempting deepfake identity-check bypasses
- The FBI said some intrusions progressed from initial targeting to privileged access in under three hours
- Censys identified 1,465 hosts with exposed ARTEX consoles as of Nov. 4
- The deepfake attempts failed server-side liveness checks in every reported attempt
Sources
- Homeland Security TodayRead ↗FBI Warns ARTEX Is Helping Hackers Automate Cyber Intrusions at ‘Machine Speed’
- cyberpressRead ↗Hackers Use ARTEX and Other AI Agents to Automate Reconnaissance, Exploitation, and Data Theft
- The RegisterRead ↗Attackers are using AI agents to automate cybercrime at 'industrial scale,' FBI and AWS warn
- news.lavx.huRead ↗Attackers used AI agents to automate cybercrime at 'industrial scale,' FBI and AWS warn
Details
Amazon Threat Intelligence detected two ARTEX campaigns through its MadPot honeypot sensors between Oct. 3 and Nov. 5, according to The Register’s report. In one, agents stole API keys and used them to seek access to AI models and related cloud services. The other targeted identity checks at two cryptocurrency exchanges with deepfake videos and forged ID images.
The deepfake operation reverse-engineered a vendor’s liveness-check protocol and positioned a face within 7 pixels of a target during its movement sequence, but the vendor’s server rejected every attempt. In a separate test, agents built a real-time deepfake pipeline on a consumer-grade GPU that ran at 11.5 frames per second. The report says the campaign also worked on synthetic camera metadata for ID checks.
The FBI said ARTEX can shorten the time needed to combine attack techniques, with some intrusions reaching privileged access in under three hours. Its alert cited August attacks on South Korean financial institutions that exposed data belonging to 67,000 people and a September supply-chain campaign involving ConnectWise Cloud Backup that compromised more than 2,250 downstream Microsoft tenants.
The ARTEX developer removed the original GitHub repository in October, but forks and a commercial version remained, the report says. Censys identified 1,465 hosts with exposed ARTEX consoles as of Nov. 4; the FBI said the total number of deployments is higher when private and firewalled systems are included.
Written from coverage as of
Images
- cyberpress
- news.lavx.hu
Stocks in this story
Change since open: 20 Nov to 21 Nov 2026
Last refresh