SentinelOne Executive Says Daybreak Cut Malware Analysis to Eight Hours

SentinelOne Chief AI Officer Gregor Stewart described a malware reverse-engineering task completed in roughly eight hours with OpenAI’s Daybreak models, compared with four to six weeks for a specialist, a difference the company says could expand threat research capacity.

Stewart’s example appeared in an account published Oct. 14, 2026. He said the faster analysis could let SentinelOne researchers investigate more threats and look for shared components and relationships among them. He also emphasized that human accountability remains part of the workflow.

The example follows SentinelOne’s Sept. 3 expansion of its Wayfinder Frontier AI Services with OpenAI models, beginning with GPT-5.6-Cyber. The related capabilities, which included code-risk analysis, compromise assessments and support for malware analysis, were in private preview at the time.