FBI says AI hacking tool ARTEX can break into networks in under three hours | #hacking | #cybersecurity | #infosec | #comptia | #pentest | #hacker

The FBI has warned that hackers are using ARTEX, an AI hacking tool, against at least ten sectors. They range from banks and defence to healthcare and law. Some intrusions went from first targeting to privileged access in under three hours, the bureau said.
The warning came in a FLASH alert published on Friday, 13 November. It says the tool lets attackers with little skill run complex attacks on many targets at once.
ARTEX is an open-source framework in which AI agents carry out each stage of an attack with little human input. It has no AI model of its own. It plugs into models from Anthropic and OpenAI as well as Chinese-origin ones. DeepSeek’s Flash models are the most used on attackers’ servers, the FBI said. They are cheap and show no visible guardrails when attackers present a request as penetration testing.
The alert ties the tool to the theft of personal data on about 67,000 people from South Korean banks. It also ties it to the ConnectWise supply chain attack, which reached more than 2,250 downstream tenants.
Pulled from GitHub, still spreading
A developer called “Autumn-27” released ARTEX on GitHub in July. It gathered an estimated 1,500 to 2,000 stars within ten weeks. The developer said the tool was for authorised testing only. After reports of abuse, the repository came down on 8 October.
Forks, mirrors, archived copies and a paid version called ARTEX PRO keep it in circulation, the FBI said. Researchers counted more than 1,600 ARTEX consoles exposed online in early November. The real number of deployments is likely much higher, the bureau added.
What Amazon’s decoys caught
Amazon published its own findings on the same day. Its MadPot decoy sensors look like popular AI model services. Between 3 October and 5 November, they received requests from ARTEX agents that carried the agents’ own records of their campaigns. Amazon’s chief information security officer, CJ Moses, set this out in a company blog post.
In the largest campaign, agents stole API keys and other credentials for AI models. The stolen keys then paid for the AI that powered the hunt for more keys. Targets included AI gateways such as LiteLLM and app builders such as Dify and Ollama, The Register reported.
A second campaign tried to beat identity checks at two cryptocurrency exchanges with live deepfake video and forged ID photos. The verification vendor’s server rejected the deepfake every time, Moses wrote.

“The agents succeeded where basic security hygiene was missing, including default or hardcoded credentials, keys left in code and public artifacts, unauthenticated interfaces and inference endpoints, and unpatched software,” Moses wrote.

Barracuda changes every customer credential
Security company Barracuda has also disclosed an ARTEX-linked breach, TechCrunch reported. The tool found an exposed credential for a third-party service that Barracuda had not revoked when the work ended. That gave the attackers a way into stores of credentials Barracuda uses to collect data from its customers’ systems.
Its logs could not rule out access, so Barracuda changed all its customer credentials, according to TechCrunch. Its chief information security officer, Arve Kjoelen, wrote:

“Defenders who rely on spotting malicious activity and then responding to it will often arrive after the attacker has finished.”

Google also named ARTEX among several AI attack frameworks in its latest threat report on Monday. The FBI asked organisations to list every internet-facing system, including partner portals and employee apps. It also urged them to keep admin interfaces off the open internet.


Click Here For The Original Source.

——————————————————–

............