Adaptive detonation technology has been launched by Joe Security to help analysts uncover malicious behavior that is often concealed by sophisticated evasion techniques, as first reported by CSO Online.
The new system utilizes artificial intelligence agents to monitor execution results within a sandbox environment. When a malware sample fails to reveal its malicious intent due to obstacles like timing checks, firmware-based rejection, or missing dependencies, the AI intervenes. These interventions include patching blocking checks, creating necessary files, adjusting environmental settings, or providing missing command-line arguments. Once these modifications are applied, the sandbox re-analyzes the sample to verify if the hidden activity is exposed.
This process provides security teams with deeper visibility into evasive threats that might otherwise remain dormant during initial analysis. By automating the identification of unmet execution conditions, the tool reduces the manual effort required for reverse engineering. Adaptive detonation is currently available as a plugin for Joe Sandbox Cloud. This development reflects a broader industry shift toward using agentic technology to combat increasingly complex malware that is designed to detect and bypass traditional automated analysis environments.
Source: CSO Online
