The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.

Darktrace stated that LockBit’s claims of compromising its internal systems were unfounded following a full security review. The company reported no evidence of a breach and confirmed its services remained fully operational with no customer impact. Investigations are ongoing, but Darktrace expressed confidence in the security of its systems and data protection measures.

Darktrace announced its selection as one of the first cybersecurity firms to integrate behavioral risk signals into Microsoft Agent 365. The integration will use Darktrace / SECURE AI™ to generate organization-specific signals based on each customer’s Unique Behavioral Profile. These signals will help security teams assess whether AI agents’ actions align with expected behavior within the Agent 365 portal.

The integration aims to address growing concerns around AI agent security by providing visibility into agent identities, permissions, and actual behavior. Darktrace’s platform will surface risk signals directly within Agent 365, allowing customers to view Microsoft and Darktrace data in a single interface. This approach helps identify potential compromises, misconfigurations, or deviations from intended agent activity.

Darktrace emphasized the importance of understanding agent behavior as AI adoption expands across enterprises. Ed Jennings, CEO, highlighted the need for security teams to monitor not just permissions but also real-world agent actions. Microsoft’s Nirav Shah underscored the integration’s role in delivering a unified view of risk signals within the Agent 365 ecosystem.