Technology news and discussion on TalkFuse.
Attackers are using AI agents to automate cybercrime at ‘industrial scale,’ FBI and AWS warn
The FBI and Amazon Threat Intelligence say attackers are using the open-source ARTEX framework to automate cybercrime against organisations in critical sectors. Two campaigns show how AI agents can speed up attacks and coordinate several tasks at once, although the underlying techniques are not new.
Amazon sensors detected the campaigns between 3 October and 5 November. One used stolen AI credentials to obtain access to models; another tried deepfakes and forged identity documents to bypass checks at two cryptocurrency exchanges, but the deepfake attempts were rejected. The FBI said some intrusions reached privileged access in under three hours, citing attacks that affected about 67,000 people in South Korea and more than 2,250 Microsoft tenants. AWS said basic security weaknesses, including exposed keys and unpatched software, enabled the attacks.
In short
- ARTEX agents automated attacks on critical organisations.
- Some intrusions reached privileged access in under three hours.
- Exposed keys and weak security helped attackers succeed.
Explained simply
Cybercrime is the use of computers and online services to steal information, disrupt organisations or gain access to their systems. Critical sectors include services such as healthcare, energy and finance, where disruption can affect many people. Attackers may combine stolen login details with software that searches for weaknesses in computer networks.
ARTEX is an open-source framework: its code is publicly available and can be adapted by different users. AI agents are software systems that can carry out connected tasks with limited human input, such as gathering information or helping operate other tools. Using them can speed up parts of an attack, even when the techniques involved are familiar.
The FBI investigates federal crimes in the United States, while Amazon Web Services (AWS) provides cloud computing and monitors threats to its customers. Their account matters because it describes how quickly attackers may move through poorly protected systems, and how weaknesses such as exposed access keys or out-of-date software can put organisations and their users at risk.