Abnormal AI launched AI Cloud Security this week. The product extends its behavioral AI engine to cloud environments. It spots risky or malicious behavior from AI agents inside customer systems. OpenAI Daybreak models support investigation and response. The approach is described as fighting bad AI with good AI. The news came with OpenAI's Cyber Summit. The tool is part of the Daybreak Defense Network. It is in private preview for select Abnormal customers. General availability is set for the fourth quarter of 2026. The system builds a living model of every identity in the cloud. This covers human users, service accounts, API keys, and AI agents. It learns normal behavior for each. Unusual activity can signal an AI-driven breach. Customers choose automatic action or human review based on severity. Investigation features use OpenAI Daybreak models to help teams examine the anomalies found by the behavioral engine.
Abnormal AI launched AI Cloud Security this week. The product extends its behavioral AI engine to cloud environments. It spots risky or malicious behavior from AI agents inside customer systems. OpenAI Daybreak models support investigation and response. The approach is described as fighting bad AI with good AI. The news came with OpenAI's Cyber Summit. The tool is part of the Daybreak Defense Network. It is in private preview for select Abnormal customers. General availability is set for the fourth quarter of 2026. The system builds a living model of every identity in the cloud. This covers human users, service accounts, API keys, and AI agents. It learns normal behavior for each. Unusual activity can signal an AI-driven breach. Customers choose automatic action or human review based on severity. Investigation features use OpenAI Daybreak models to help teams examine the anomalies found by the behavioral engine.