Theori Whitepaper Flags Gaps in ARTEX AI Security Tool
Theori published a whitepaper on the AI security tool ARTEX and released free detection rulesets for Suricata, Sigma, YARA and ModSecurity on GitHub.
Theori announced on October 15, 2026 that it published a whitepaper analyzing the AI security tool ARTEX and released related detection rulesets free of charge through its GitHub repository. The announcement was reported by Digital Today.
The whitepaper examines AI-based attack tools whose use has increased rapidly in recent months, according to the company. Theori's analysis found that these tools leave behind identifiable information and behavioral patterns, including request order, frequency and repetition.
The company also described a safeguard verification experiment on ARTEX in which test data was actually deleted during a simple vulnerability scan. According to Theori, the LLM-based safeguard did not block the action even when it was enabled.
Alongside the whitepaper, Theori published detection rulesets for Suricata, Sigma, YARA and ModSecurity, making them available at no cost via GitHub.
Since late September 2026, breaches targeting domestic financial institutions in South Korea have occurred, in which the open source AI security tool ARTEX was used.
Quick answers
What did Theori release for free?
Theori released detection rulesets for Suricata, Sigma, YARA and ModSecurity for free through its GitHub repository, alongside a whitepaper on the AI security tool ARTEX.
What did Theori's experiment on ARTEX find?
In Theori's safeguard verification experiment, test data was actually deleted during a simple vulnerability scan, and the LLM-based safeguard did not block it even when enabled.
What patterns do AI-based attack tools leave behind?
According to Theori's analysis, AI security tools leave identifiable information and behavioral patterns such as request order, frequency and repetition.