Canada warns servers running AI hacking tool ARTEX target its organizations
The Canadian Centre for Cyber Security says AI-enabled offensive security frameworks are spreading, and that servers running ARTEX have targeted Canadian organizations for months.
Key takeaways
- Canada's Cyber Centre says attackers increasingly use AI-enabled hacking frameworks to find and attack internet-facing systems.
- About 46% of the AI-enabled offensive security infrastructure it identified was tied to ARTEX, 28% to CyberStrikeAI and 14% to PentAGI.
- Servers running ARTEX have targeted Canadian organizations for several months; the agency has not named an actor and says targeting is not proof of compromise.
- ARTEX stopped being open source on October 8, after it had been copied more than 12,000 times, the agency says.
- The advisory urges suppressing detailed error messages, watching cloud identity logs and patching internet-facing systems quickly.
The Canadian Centre for Cyber Security has warned Canadian organizations that attackers are making growing use of automated and AI-enabled hacking frameworks to find and attack systems exposed to the internet. In a new advisory, the agency names ARTEX, CyberStrikeAI and PentAGI among a rising number of such frameworks it is tracking.
Of the AI-enabled offensive security infrastructure the Cyber Centre has identified worldwide, about 46% was linked to ARTEX, 28% to CyberStrikeAI and 14% to PentAGI, according to the advisory, with other frameworks making up the rest.
Are Canadian organizations being hit?
The agency says it is aware of activity aimed at Canadian organizations in multiple sectors from servers running ARTEX, which it says has gone on for several months and is continuing. It has not attributed this activity to a specific actor, and it stresses that being targeted is not proof of a breach; it shows these attack frameworks can reach systems in Canada. The Cyber Centre says it notifies organizations when it learns they are being targeted.
The advisory also points to public reporting that has linked ARTEX to state-aligned activity, citing a Microsoft report that a PRC actor it tracks as Storm-0940 used the framework against hundreds of organizations. Other reported ARTEX activity remains unattributed, the Cyber Centre notes.
What is ARTEX?
ARTEX is an autonomous penetration testing framework that uses AI agents for steps from reconnaissance to data theft, BleepingComputer reported when the FBI issued its own warning about the framework last week. According to the Cyber Centre, ARTEX stopped being open source on October 8, after it had already been copied more than 12,000 times. The FBI said the framework is still available through forks, archived copies and a commercial edition, and that researchers had found more than 1,600 publicly exposed ARTEX consoles as of early November, BleepingComputer reported.
How do these attack frameworks change an attack?
The Cyber Centre says attacks are not yet fully autonomous: people still pick the targets, set goals and oversee the work, but machines now do more of the steps in between, leaving defenders less time. In some cases it previously described, attackers went from picking a target to privileged access in cloud systems in less than three hours, the agency says.
Because these attack frameworks work through trial and error at high volume, much of this activity is currently noisy enough for capable defenders to spot, the advisory says. It expects such activity will likely grow more targeted and harder to spot as the tools mature.
What should organizations do?
The agency recommends phishing-resistant multi-factor authentication, quick patching of internet-facing systems and tighter limits on what users, apps and service accounts can access. Specific to these attack frameworks, it advises turning off detailed error messages in production, since attackers' automated agents use them to repair a failed exploit and retry in seconds. It also suggests watching cloud identity logs for new secrets or permissions added to applications, and checking for unusual outbound connections from servers.
Organizations that clean up after an incident should re-scan their external systems, rotate every credential the attacker could have reached and look for persistence that survives a password change, the Cyber Centre says.
Questions readers are asking
Which AI cyberattack frameworks does the advisory name?
ARTEX, CyberStrikeAI and PentAGI, which together account for most of the AI-enabled offensive security infrastructure the Cyber Centre has identified.
Have Canadian organizations been breached with ARTEX?
The Cyber Centre says servers running ARTEX have targeted Canadian organizations for months, but that targeting is not an indication of compromise. It has not attributed the activity to a specific actor.
Why does the agency advise hiding error messages?
It says attackers' agents use detailed errors and stack traces to repair a failed exploit and retry in seconds, so suppressing them weakens these attacks.
Sources
- #artex
- #canada
- #cyber-centre
- #ai-attacks
- #threat-advisory
Get the day's top AI & tech stories on WhatsApp: Follow the NextFish channel
Get stories like this every morning.
Related
Cybersecurity3 min
Amazon says hackers used AI tool Cairn to find and exploit a zero-day
Amazon Threat Intelligence says AI agents found an unknown flaw in a medical imaging app and used it to take over a university server in East Asia, then breach a second university.
Cybersecurity2 min
Supplier hack exposes payee data at more than 40 Irish credit unions
Belfast-based Wellington IT's mobile banking platform was breached in early October; credit unions were told in mid-November. Officials see no evidence funds were accessed.
Cybersecurity2 min
Chase, Visa and Mastercard confirm incidents at outside vendors
Chase says some customer names and emails may be affected. A ransomware gang called n0n has claimed the attack and leaked data it says came from Chase.
Cybersecurity2 min
F5 issues emergency fixes for critical BIG-IP remote code execution flaws
The bugs can crash BIG-IP's traffic engine and may let unauthenticated attackers run code. F5 lists no mitigations besides the hotfixes.