VMTech
Discuss a project →

FBI warns Artex AI tool can compromise networks in under three hours

FBI warns Artex AI tool can compromise networks in under three hours

The FBI has warned that the open-source AI-powered penetration-testing platform Artex has been used in cyberattacks against organizations in the United States and internationally, with some intrusions progressing from initial targeting to active compromise in under three hours. The advisory names financial services, government, technology, defense, telecommunications, internet providers and law firms among the sectors targeted in recent months.

Artex is designed to probe a network’s defenses and can be connected to hosted frontier large language models from Anthropic and OpenAI, as well as open-source and non-US models. The FBI said the tool can generate substantial traffic and data, making its activity potentially noisy enough to trigger defensive alerts. Its speed, however, can make it difficult for human-led response processes to keep pace.

Campaigns affected financial institutions and service providers

Since Artex first became publicly available in July, the FBI says it has appeared in several attacks. In South Korea, attackers used it against financial institutions where weak security defenses and missing multi-factor authentication contributed to the compromise. The incidents resulted in the theft of personal data belonging to at least 68,000 people.

The FBI also confirmed Microsoft’s finding that attackers used Artex against IT management company ConnectWise and obtained access to Microsoft 365 cloud environments belonging to at least 2,250 customers. Microsoft assessed the attackers as likely Chinese, and said data was stolen from affected ConnectWise customers.

The scale of this activity aligns with ARTEX campaigns against more than 200 organizations reporting on AI-enabled ARTEX campaigns against more than 200 organizations, while the FBI’s advisory adds detail on the sectors, victims and attack paths involved.

Exposed credentials remain a central entry point

The FBI said Artex can quickly scan a target network for weaknesses, including weak protections, exposed passwords and stolen credentials. Once attackers establish a foothold, the platform can help them move through the environment in search of databases, mailboxes and other sensitive information.

Barracuda disclosed an Artex-linked breach by financially motivated criminals after the tool identified an exposed credential for a third-party service. The credential had not been revoked after the vendor relationship ended. The access enabled attackers to reach stores of credentials Barracuda used to collect data from customer systems.

Barracuda said its logs were not sufficiently detailed to rule out access and it reset all customer credentials on the assumption that compromise was possible. Its chief information security officer, Arve Kjoelen, said defenders who wait to spot malicious activity and then respond may arrive after an attacker has completed the operation.

Accessible software creates a continuing risk

Artex’s developer made the software closed-source after it was used in attacks, but the FBI said the previous open-source version continues to circulate and remains widely accessible. Researchers also found hundreds of publicly exposed Artex systems online without password protection, apparently due to mistakes or misconfiguration; this included the system used to target South Korea.

For businesses, the practical implication is to reduce the entry points the FBI identified: require multi-factor authentication, revoke credentials when third-party access ends, protect externally exposed systems and investigate anomalous high-volume activity quickly enough to limit automated lateral movement.

#cybersecurity#artificialintelligence#identitysecurity#threatdetection
Open analytics
On the site 0 views
min read 4 16.11.2026
Instagram

FBI warns Artex AI tool can compromise networks in under three hours

Open the post on Instagram ↗