ShieldForce Colored Variance Logo
ARTEX: The AI Hacking Tool the FBI Says Can Breach a Network in Under 3 Hours
AI Security

ARTEX: The AI Hacking Tool the FBI Says Can Breach a Network in Under 3 Hours

ARTEX is a free, open-source AI tool that uses teams of AI agents to run a cyberattack from start to finish. It was built for security testing, and attackers now use it for real intrusions.

By Obi Ibeto, Founder & CEO, ShieldForce · Last updated November 19, 2026 · For owners, administrators, and executives in home health, hospice, palliative care, nursing homes, and FQHCs

Key takeaways

  • ARTEX is a free, open-source AI tool that uses teams of AI agents to run a cyberattack from start to finish. It was built for security testing, and attackers now use it for real intrusions.

  • In a FLASH alert dated November 13, 2026, the FBI said some ARTEX intrusions went from first targeting to administrator-level access in under three hours.

  • Healthcare is one of ten sectors the FBI names as targeted, along with financial services, government, and defense.

  • ARTEX gets in through the systems that get the least attention: vendor and partner portals, old software, staff-facing tools, and leaked passwords.

  • Because the attack takes hours, the defense has to work in minutes. That means fewer exposed systems, tight vendor access, and 24/7 monitoring with someone ready to act.

Until this year, breaking into a network took a skilled person days or weeks of patient work. The FBI now says an AI tool can do most of that work on its own, and finish before lunch.

The tool is called ARTEX. If you run a home health agency, hospice, nursing home, palliative care program, or community health center, here is what it is, what the FBI found, and what to do about it.

What is ARTEX?

ARTEX is an open-source penetration testing framework that uses several AI agents to carry out the steps of a cyberattack with very little human help. Penetration testing means a planned, authorized attack on your own systems to find weak points before criminals do. ARTEX automates that work, and attackers have turned it on real targets.

A developer known as "Autumn-27" released ARTEX on GitHub, a public code-sharing site, in July 2026. The developer removed it on October 8, 2026 after reports of abuse and said it was only ever meant for authorized testing. Copies and spin-off versions are still easy to find.

ARTEX has no AI model of its own. It plugs into existing ones. According to the FBI, attackers most often pair it with low-cost Chinese-built models, because those models are capable, cheap, and tend not to refuse a request framed as security testing.

What did the FBI say about ARTEX?

The FBI published a FLASH alert on November 13, 2026 warning that criminals and government-linked hackers are using ARTEX to automate intrusions worldwide. Its main findings:

  • Active since August 2026. A variety of attackers have used ARTEX against organizations around the world.

  • Ten targeted sectors. Financial services, government, technology, defense, semiconductor, energy, healthcare, manufacturing, telecommunications, and legal.

  • Under three hours. Some intrusions went from first targeting to privileged access (administrator-level control) in less than three hours.

  • Lower skill needed. People with little technical expertise can now run complex attacks on many targets at once.

  • Still spreading. Researchers counted more than 1,600 ARTEX control panels visible on the open internet by early November 2026.

The FBI did not tie ARTEX to one group. Many unrelated attackers use it.

How does an ARTEX attack work?

ARTEX strings together steps that used to need a skilled person at each stage:

  1. Scan. AI agents map everything an organization has facing the internet, including systems the organization may have forgotten.

  2. Find the weak spot. The FBI says ARTEX is especially good at finding overlooked systems: partner and supplier portals, staff and sales-support tools, mobile app back ends, old software, and third-party connections.

  3. Get in. It exploits a flaw or uses a leaked or guessed password.

  4. Climb. It chains techniques together until it has administrator-level access.

  5. Spread or steal. It copies data out, or uses one compromised vendor as a bridge into that vendor's customers.

ARTEX is noisy. It generates a lot of automated traffic, which can set off security alerts. The problem is the clock. If nobody is watching the alerts at 2 a.m., the attack is over before anyone reads them.

What attacks have been linked to ARTEX?

The FBI's alert points to two cases.

South Korean banks. Between late September and early October 2026, attackers used ARTEX against South Korean financial institutions, including Shinhan Bank, KB Kookmin Bank, and Hana Bank. The FBI says personal information from about 67,000 people was stolen. Investigators found the attackers did not break into the banks' core systems. They came in through less protected staff-facing and sales-support systems.

The ConnectWise Cloud Backup supply chain attack. In October 2026, Microsoft reported that a group it links to China compromised a backup service that IT providers use for their customers, then used ARTEX to reach toward more than 2,250 downstream customer environments. A supply chain attack is one where the attacker breaks into a vendor to get to that vendor's customers.

Both cases carry the same lesson for healthcare: the attacker came through a side door. We saw the same pattern in the UnitedHealth breach and its impact on home health agencies.

Is healthcare a target of ARTEX attacks?

Yes. The FBI lists healthcare among the ten sectors that attackers have targeted with ARTEX since August 2026. As of November 19, 2026, we have not found a publicly named home health, hospice, nursing home, or FQHC victim. The FBI's sector list is reason enough to prepare.

Why are home health, hospice, nursing homes, and FQHCs exposed?

ARTEX is built to find the systems nobody is watching. Small and midsized healthcare organizations have a lot of those.

  • Many side doors. A typical agency runs an EHR, a billing system, a scheduling or visit-verification app, a patient or family portal, remote access for field staff, and email. Each one faces the internet. See how the major platforms compare in our guides to FQHC EHR cybersecurity and securing PointClickCare.

  • Heavy vendor dependence. Your IT provider, billing company, backup service, and EHR vendor all hold access to your systems. The ConnectWise case shows that an attacker who gets into one vendor can reach thousands of that vendor's customers.

  • No one watching overnight. Most agencies have no security staff on nights and weekends. An attack that takes three hours fits easily between Friday evening and Monday morning.

  • Old accounts and passwords. Former employees, former vendors, and shared logins leave working credentials behind. ARTEX looks for exactly these. Barracuda, a security company, disclosed that an ARTEX-linked attack got in through a third-party credential it had not turned off.

  • Patient data has high value. A break-in that reaches protected health information (PHI) becomes a HIPAA breach, with notification deadlines, real costs, and possible penalties. Our FQHC ransomware case study shows how quickly that unfolds.

ARTEX also changes who can attack you. Skilled hackers used to pick large targets because their time was limited. An AI tool has unlimited time, so a 40-person agency is now worth scanning. This fits the trend in our 2026 home health cybersecurity threat report.

How can healthcare organizations defend against ARTEX?

The FBI's alert recommends five measures, according to published coverage of the alert. We've added what each one means for a healthcare organization.

  1. Reduce what faces the internet. List every system that can be reached from outside: portals, remote access, old test sites. Shut down what you don't use.

Need Audit-Ready HIPAA Compliance for Your Field Team?

Protect remote clinical staff, field laptops, mobile EVV apps, and EHR systems starting at $35/user/month.

  • Restrict access to sensitive applications. Put the EHR, billing, and admin consoles behind multi-factor authentication (MFA), a second proof of identity at login. FQHCs should review the 2026 MFA requirements for ePHI access.

  • Enforce least privilege. Give each person only the access their job needs. If a stolen account can't reach much, the attack stalls.

  • Monitor logins and account activity. Watch for sign-ins at odd hours, from odd places, or in rapid bursts.

  • Review vendor access and credentials regularly. Ask every vendor what access they hold, and remove accounts for vendors and staff who have left.

  • Three more steps we recommend for healthcare leaders:

    1. Get 24/7 monitoring with authority to act. Endpoint detection and response with managed detection and response (EDR and MDR) puts software on each device to spot suspicious behavior, with analysts who can isolate a device within minutes at any hour. Against a three-hour attack, a next-business-day response is too late.

    2. Patch fast. Apply security updates on anything internet-facing within days, not months.

    3. Keep tested, separate backups. If attackers do get in, a clean backup they can't reach is what gets you back to patient care.

    These steps also line up with our 2026 HIPAA Security Rule checklist for home health.

    What should you do if you suspect an ARTEX intrusion?

    • Call your IT or security provider immediately, at any hour.

    • Isolate affected devices from the network. Don't wipe them. Investigators need the evidence.

    • Reset passwords and sign out all sessions, starting with administrator and vendor accounts.

    • Ask your vendors whether they have seen related activity.

    • Report it to the FBI at ic3.gov or through your local field office.

    • Start your HIPAA breach assessment with your compliance lead and counsel. Our HIPAA breach notification guide covers deadlines and who must be told.

    What does ARTEX mean for your HIPAA compliance program?

    The HIPAA Security Rule requires you to assess risks to PHI and address them. AI-automated attacks belong in your next risk assessment, with two questions written down: which of our systems face the internet, and which vendors can reach our data? If your team uses AI tools, also see HIPAA and AI tools for home health agencies.

    For requirements specific to your type of organization:

    Frequently asked questions

    What is ARTEX?

    ARTEX is an open-source, AI-driven penetration testing framework released in July 2026. It uses multiple AI agents to automate the stages of a cyberattack. The FBI warned on November 13, 2026 that criminals and government-linked hackers are using it for real intrusions.

    How fast can ARTEX break into a network?

    The FBI says some ARTEX intrusions went from first targeting to privileged, administrator-level access in under three hours.

    Is healthcare being targeted with ARTEX?

    Yes. Healthcare is one of ten sectors the FBI lists as targeted since August 2026. No home health, hospice, nursing home, or FQHC victim had been publicly named as of November 19, 2026.

    Who is behind ARTEX attacks?

    No single group. The FBI says a variety of attackers use it. CrowdStrike linked the South Korean bank attacks to a Chinese-speaking, financially motivated attacker, and Microsoft linked the ConnectWise supply chain attack to a group it associates with China.

    Wasn't ARTEX taken down?

    The developer removed the original from GitHub on October 8, 2026. Copies and modified versions remain widely available, so the threat continues.

    How can a small healthcare organization defend against ARTEX?

    Reduce the systems exposed to the internet, require multi-factor authentication, limit each account's access, review vendor access, patch quickly, and have 24/7 monitoring with someone able to respond within minutes.

    How ShieldForce helps

    ShieldForce protects home health agencies, hospice and palliative care providers, nursing homes, and FQHCs against attacks that move faster than any in-house team can. For a flat $35 per user per month, we provide EDR, 24/7 managed detection and response, email security, and an incident response plan, plus audit-ready HIPAA compliance deployed in 72 hours. We work alongside your existing EHR, whether that's Homecare Homebase, WellSky, MatrixCare, Axxess, or another platform. See how our pricing compares.

    If you're not sure what your organization has facing the internet, or who is watching it at 2 a.m., let's talk. We'll show you where you stand and what to fix first.

    Schedule a conversation with ShieldForce →

    Related: FaceStealer: How North Korea's Fake Video-Call Scam Puts Healthcare Executives at Risk

    Sources

    About the author

    Obi Ibeto is the Founder and CEO of ShieldForce Corporation, a healthcare-focused managed security service provider. He has over 20 years experience in Information Technology and has published peer-reviewed research in the field.

    Share this post

    Need Audit-Ready HIPAA Compliance for Your Field Team?

    Protect remote clinical staff, field laptops, mobile EVV apps, and EHR systems starting at $35/user/month.

    Topics

    #AI Security#ARTEX AI Hacking Tool
    Free Security Assessment

    Ready to Secure Your Business?

    Don't let cyber threats put your business at risk. Discover how ShieldForce protects organizations like yours - 24/7.