joseph gabriel ragoncin
news editor
DTEX has expanded its AI risk management product with new security agents to monitor AI systems in the workplace. The move targets what is described as a gap in monitoring the behavior of AI agents.
This update adds two products to DTEX’s behavioral intelligence platform: Triage Guardian Agent and Threat Hunter Agent. The software is designed to distinguish between human and AI-driven activity, track the lineage of prompts, monitor behavioral patterns, and detect signs of data breaches.
The announcement comes as companies introduce AI co-pilots and more autonomous workflows that allow them to access corporate systems and act with some authority. This change raises new security concerns for employers already managing risks associated with human insiders, especially when AI tools can move data, respond to prompts, and interact with internal systems without close supervision.
Many existing security tools can log activity, but cannot indicate whether an AI system’s behavior is consistent with its intended use. DTEX’s enhanced AI risk management product aims to address this issue by applying behavioral analytics to both employees and AI agents.
Add product
Triage Guardian Agent is being deployed as an autonomous security agent focused on separating human behavior from AI-driven actions. Track prompt lineage and behavioral signals to identify suspicious activity, including potential data breaches.
The second addition, Threat Hunter Agent, is aimed at proactive threat detection. DTEX said it uses agent workflows to assess the changing risk landscape and identify threats not yet alerted to by traditional systems.
In addition to these tools, the AI Risk Management Platform extension can discover sanctioned and unsanctioned AI tools used across your organization. Monitor prompts, responses, and data movement to identify areas at risk as your enterprise expands its use of AI applications.
Focus on security
The announcement reflects a broader cybersecurity conversation about how companies manage AI systems that are increasingly integrated into daily operations. While many organizations have implemented policies around the use of generated AI tools by their staff, security teams are also working on software agents that can perform tasks with less direct human involvement.
It changes the nature of insider risk. In established security practices, an insider threat typically refers to an employee, contractor, or partner who misuses access or makes a mistake that compromises sensitive data. AI agents add another layer because they can execute instructions at high speed and at scale, but the reason behind their output may not be fully visible to managers and analysts.
DTEX centers its latest update on this distinction, arguing that while companies are rapidly deploying AI systems, they still have limited tools to determine whether those systems are operating in accordance with their safe intentions.
operational claims
As evidence of operational value, DTEX pointed to recent implementations in government agencies. It says its customers have saved 40 hours per month per analyst and more than 500 hours per year, allowing security teams to move from reactive alert handling to more proactive threat hunting.
DTEX also said it achieved 100% accuracy on the same deployment, but did not provide details about the measurement period, scope of testing, or benchmarks used in its claim.
The focus on analysts’ time is noteworthy as security operations centers face increased alert volumes, staff shortages, and an increasing number of tools. Vendors are increasingly incorporating AI-based monitoring as a way to reduce repetitive triage tasks and free up analysts to focus on investigations that require judgment.
DTEX’s approach is based on behavioral intelligence, a category often associated with user activity monitoring and insider risk programs. The company argues that by extending its model to AI systems, the behavior of machines should be scrutinized similar to the behavior of human users who access sensitive information.
This position may be attractive to organizations in regulated sectors and government environments where data processing and monitoring of system access is already tightly controlled. It also raises practical questions about governance, such as how companies define acceptable AI behavior, how to distinguish between authorized and unauthorized use, and how to respond when AI agents deviate from expected patterns.
For buyers, much will depend on whether such systems can reduce false positives while providing analysts with enough context to understand why the AI tool behaved in a certain way. As more companies move from experimenting with AI assistants to incorporating autonomous workflows into daily operations, questions of intent are becoming a more important part of cybersecurity oversight.
