Hack The Box Launches AI Range to Benchmark Autonomous Security Agents

Hack The Box (Hack The Box) has unveiled HTB AI Range, a controlled cyber range built to test and benchmark the safety, limits, and capabilities of autonomous AI security agents, announced in a press release. The platform replicates live cyber battlegrounds, allowing enterprises to evaluate both AI agents and human operators under realistic conditions.

HTB AI Range is designed for enterprises, managed security service providers, and government agencies to stress-test AI models and measure hybrid human-machine performance. It includes thousands of continuously updated offensive and defensive targets and aligns with frameworks such as MITRE ATT&CK, NIST/NICE, and OWASP Top 10.

According to Hack The Box, the platform addresses the growing need to validate AI systems in operational contexts where human oversight remains essential. The company previously hosted an AI vs. Human Capture The Flag event, where autonomous AI teams solved 95% of easy-tier challenges but fell short on multi-step problems compared to human teams.

Hack The Box stated that AI Range aims to prepare organizations for a hybrid defense future, where humans and AI systems operate together to counter increasingly automated cyber threats.