OpenAI unveils plan to protect critical services from AI cyberattacks

Mountain View, California - Cybersecurity firm SentinelOne has broadened its AI services by incorporating OpenAI's Daybreak models into its Wayfinder Frontier offerings. The company will utilize OpenAI's GPT-5.6-Cyber model to analyze code vulnerabilities, identify compromise risks, and examine malware. SentinelOne's AI-powered code risk analysis scans customer code repositories for security flaws, implants, exposed secrets, and supply-chain threats.

Meanwhile, the AI-enabled compromise assessment evaluates telemetry against detection rules to pinpoint gaps in security posture, such as risky use of VPNs, proxies, or remote-management tools. SentinelOne's research division, SentinelLABS, evaluated OpenAI's GPT-5.6-Cyber model and found it to perform well in malware reverse engineering and analysis.

The new model will complement SentinelOne's existing security toolset to tackle specific threats. Steve Stone, SentinelOne's Chief Customer Officer, emphasized that attackers are leveraging AI to discover and exploit weaknesses more swiftly and widely, and that the company's role is to help close this gap by providing elite experts with access to the most advanced AI models available.

The services combine AI models with SentinelOne's security analysts to deliver customers prioritized threat assessments and remediation guidance. When scans identify potentially malicious samples, AI-assisted workflows aid in disassembly and deobfuscation, furnishing assessments of the sample's capabilities, persistence, and command-and-control mechanisms.

Wayfinder Frontier AI Services is now available, with the expanded model lineup featuring OpenAI's Daybreak models being rolled out in a private preview, with broader availability planned for later.