SentinelOne Launches GPT-5.6-Cyber for Enhanced Malware Analysis

SentinelOne Inc. has expanded its Wayfinder Frontier AI Services by integrating GPT-5.6-Cyber, a model optimized for malware reverse engineering. This decision follows benchmarking by SentinelLABS, which identified GPT-5.6-Cyber as the top performer in analyzing military-grade malware. The new capabilities include scanning customer code repositories for OWASP-class flaws, exposed secrets, and supply-chain risks, with findings mapped to the MITRE ATT&CK framework. Offensive-security analysts at SentinelOne validate all results before they are shared with customers. The compromise assessment feature evaluates telemetry against detection rules to identify security posture gaps, particularly concerning the use of VPNs and remote management tools. This rollout is part of a broader strategy to counteract the increasing use of AI by attackers to exploit vulnerabilities. The launch occurred on September 4, 2026.

Key Points: • SentinelOne launched GPT-5.6-Cyber for advanced malware analysis. • The model was benchmarked as best-in-class for military-grade malware reverse engineering. • New capabilities include scanning for OWASP flaws and evaluating security posture gaps.