Hacker touching lock showing AI hacking tool for cyber attacks

FBI Warns Cybercriminals Are Weaponizing ARTEX AI Hacking Tool to Accelerate and Scale Cyber Attacks

The Federal Bureau of Investigation warns that ARTEX, an open-source AI hacking tool, is powering a new wave of cyber attacks by lowering the entry barrier and reducing the time required to breach organizations, including critical infrastructure, to just under three hours.

“The framework enables threat actors with minimal technical expertise to conduct complex, multi-target operations simultaneously, fundamentally accelerating attack velocity and expanding operational scale,” the agency stated.

Malicious actors repurpose AI hacking tool ARTEX for cyber attacks

ARTEX is a legitimate open-source penetration testing framework that cybercriminals have repurposed and added to their arsenal to target organizations. It translates natural-language instructions into actionable machine commands, executed with minimal human interaction, automating cyber attacks. ARTEX also integrates easily with various frontier AI models, from OpenAI’s ChatGPT and Anthropic’s Claude to open-source and Chinese LLMs, such as Grok, DeepSeek, GLM, and Kimi.

However, Chinese AI models are most popular among cybercriminals due to their capabilities, relatively low costs, and fewer safety guardrails when used for pentesting or cybercrime, with DeepSeek Flash being the most exploited.

ARTEX released its open-source model on GitHub in July 2026 but removed it by October 8, 2026, after learning that cybercriminals were abusing it to target organizations. However, many malicious actors had already downloaded the AI hacking tool and were using it in state-sponsored or financially motivated cyber attacks. The AI model is also frequently shared on code-sharing platforms, making it widely available.

“However, the framework continues to proliferate through community forks, localized variations, archived copies, and a commercial version (ARTEX PRO), ensuring continued availability to threat actors,” the FBI warned.

By November 2026, over 1,600 public deployments of the AI hacking tool existed, with many more tucked away behind firewalls. Additionally, some publicly accessible ARTEX instances could be shared by multiple users, potentially fueling more cyber attacks. According to threat intelligence firm Censys, some ARTEX deployments also hosted other AI attack frameworks.

“Information security teams have always raced the clock for vulnerability detection and remediation. With the public release of ARTEX, threat actors have changed the tempo,” said Morey J. Haber, Chief Security Advisor at BeyondTrust.

While ARTEX produces high volumes of “forensically noisy traffic,” the attack velocity prevents defenders from reacting in time.

According to the federal law enforcement agency, the AI coding tool exploits various vulnerabilities and security weaknesses, including XML External Entity (XXE) flaws, compromised third-party vendor credentials, hard-coded administrator passwords, missing API authentication, remote code execution, SQL injections, VPN misconfigurations, Server-Side Request Forgery (SSRF), third-party access, and cross-account cloud trust relationships.

Its ability to scan multiple targets, identify security vulnerabilities, and chain exploits makes ARTEX a powerful tool for penetration testing and hacking.

“What stands out here is how little it now takes to run a full intrusion,” said Bri Frost, Director of Product Management, Cloud Range. “Someone with minimal expertise can give an agent an objective and let it move from reconnaissance to privileged access in a matter of hours. AI frameworks and the new AI swarm attacks are going to force incident response to change. You can’t just mitigate one vulnerability and put the system back online.”

Meanwhile, the FBI advises defenders to implement the recommended mitigations and report suspected ARTEX cyberattacks.

AI hacking tools are frequently used in cyber attacks

Cybercriminals have exploited the AI hacking tool to target South Korean banks and steal data from over 67,000 customers.

Attackers also abused the AI hacking tool to breach ConnectWise, an IT management software company, targeting over 2,250 downstream customers.

According to the FBI advisory, attackers have abused the AI hacking tool to target numerous critical infrastructure organizations, including financial services, government, technology, defense, semiconductor, energy, healthcare, manufacturing, telecommunications, and legal industries.

State-sponsored and low-skilled cyber actors also use other legitimate AI tools, such as the Cairn, CyberStrikeAI, HexStrike AI, and PentAGI penetration-testing frameworks, to target organizations, according to a Google Threat Intelligence report.

“Check Point Research estimated that one two-week ARTEX campaign against ten targets cost $700 to $1,500 in model fees,” said Collin Hogue-Spears, Senior Director of Solution Management at Black Duck. “Agents that Amazon ties to ARTEX logged feeding stolen AI model API keys back into the framework to run the next key search. The FBI’s request to ARTEX victims includes token consumption and API billing anomalies, so investigate an unexplained jump in AI model spending for stolen keys.”