Malware · Threat Intelligence
FBI warns Artex AI tool lets attackers automate cyberattacks at speed
The FBI has warned that the Artex AI tool is being used by criminals and state actors to automate complex cyberattacks, cutting intrusion times to under three hours.
Ross Grant2 min read

The short version
The FBI says a tool called Artex is helping attackers use artificial intelligence to break into computer systems much faster than before. This tool can automate many steps in a cyberattack, so even people with little technical skill can use it. It has already been used in attacks on banks and government agencies. The original creator removed it from the internet, but copies are still spreading.
By the numbers
- 1,600+publicly exposed Artex console instances found by CensysRecorded Future News
- under 3 hourstime from targeting to privileged access in some attacksRecorded Future News
What happened?
The FBI has issued a flash alert warning that the Artex AI tool is being used by cybercriminals and nation-state actors to automate cyberattacks. The tool can reduce the time from initial targeting to privileged access to under three hours. It has been used against government agencies and critical sectors, including recent attacks on financial institutions in South Korea and Japan. The developer has removed the original code from GitHub, but the tool remains widely available.
How the Artex tool works
The FBI has issued a flash alert about Artex, an open-source AI platform that automates multiple stages of cyberattacks. According to Recorded Future News (opens in a new tab), Artex can turn any large language model (LLM) into a hacking assistant. It translates high-level instructions into machine commands, allowing attackers to move from initial targeting to privileged access in under three hours. The tool can automate reconnaissance (gathering information about targets), exploitation, and data theft with minimal human input after the initial setup.
Artex supports integration with a wide range of AI models, including well-known ones like OpenAI's GPT and Anthropic's Claude, as well as Chinese-developed models such as DeepSeek, GLM, Grok and Kimi. The FBI noted that Chinese models are most common in attacker-controlled Artex instances, especially the DeepSeek Flash family, which is popular for its low cost and lack of security restrictions.
Who is affected
The FBI advisory says Artex has been used against government agencies and dozens of critical sectors worldwide. Targets include financial services, technology, defence, semiconductors, energy, healthcare, manufacturing, telecommunications and legal industries. Recent attacks linked to Artex have hit South Korean and Japanese financial institutions, causing national scandals. Microsoft reported that Chinese state-backed actors used Artex in a supply chain attack on IT management software company ConnectWise. Amazon found that Chinese-linked groups used Artex to steal credentials for accessing AI models and to automate further attacks.
The response
Artex was released in August on GitHub by a developer identified as Li Puhua, a Chinese cybersecurity engineer. After reports of abuse, Li removed the repository and denied responsibility for criminal use. Despite this, the FBI says Artex remains widely available through forks (copies), archived versions and a commercial edition. Research from Censys found over 1,600 publicly exposed Artex consoles, with the real number likely higher due to private deployments.
What you should do
The FBI and experts warn that Artex can generate a high volume of attack attempts, which may trigger security alerts but at a speed that can overwhelm defenders. Security teams should review their detection and response processes to handle automated, multi-stage attacks. Special attention should be paid to less-monitored systems like partner portals, mobile backends, legacy applications and third-party integrations, as Artex is adept at finding and exploiting these weak points.
At the time of writing, only Recorded Future News (opens in a new tab) has reported on the FBI alert. The full advisory has not been published publicly, and details on the total number of victims or financial losses are not available.
Why it matters
Artex lowers the barrier for launching complex cyberattacks, making it easier for more attackers to target organisations at scale. Its speed and automation challenge existing security defences, especially for less-protected systems and third-party connections.
Key points
- Artex automates cyberattacks, cutting intrusion times to under three hours.
- The tool works with many AI models, including popular Chinese and Western LLMs.
- Critical sectors and government agencies have been targeted, including in South Korea and Japan.
- The original code was removed, but Artex remains widely available through copies and commercial versions.
- Security teams should watch for high-speed, automated attacks, especially on less-protected systems.
Sources
Recorded Future News
Reference


