DTEX has enhanced its AI Risk Management platform with new capabilities designed to help organisations manage the growing security risks associated with generative AI tools, copilots and autonomous AI agents.
As AI technologies become increasingly integrated into enterprise operations, they are gaining access to sensitive data, business systems and critical workflows. DTEX says traditional security solutions can often monitor AI activity but struggle to determine whether actions are aligned with intended behaviour or introduce potential risk.
According to DTEX CEO Marshall Heilman, AI agents are rapidly evolving into active participants within enterprise environments, capable of accessing information, interacting with systems and making autonomous decisions.
The expanded platform uses behavioural intelligence to identify and mitigate risks originating from both human users and AI systems. By combining AI risk management with automated investigation and response capabilities, organisations gain greater visibility and control as AI adoption accelerates.
DTEX notes that AI agents represent a new category of enterprise risk because they can interpret instructions, access sensitive information, interact with external platforms and make decisions with limited human oversight. The company believes effective protection requires understanding not only what actions are taking place, but also the intent behind those actions and how behaviour changes over time.
The platform provides visibility into both approved and unauthorised AI usage across enterprise environments. It can identify AI activity across users, endpoints, browsers, applications, development environments and embedded AI tools, while also detecting shadow AI deployments and unmanaged copilots in real time.
Additional functionality includes monitoring prompts, responses, uploads, downloads and AI-generated content to identify potential exposure of source code, intellectual property and sensitive business information. The system also classifies AI interactions for compliance, auditing and threat investigation purposes while applying behavioural analysis to better understand user and agent intent.
DTEX has also introduced capabilities to distinguish between human and AI-driven activities, including visibility into Computer Use AI (CUI) systems. This enables organisations to track agent instructions, task execution methods and actions performed across enterprise systems. The platform is designed to detect and prevent autonomous AI-driven data exfiltration through a combination of behavioural monitoring, prompt lineage analysis and AI-specific risk models.
In one early deployment, DTEX identified an autonomous AI agent that was exposing sensitive enterprise information despite operating within approved permissions and workflows. The company said behavioural analysis and prompt lineage tracking enabled the issue to be detected before it resulted in a security breach.
Alongside the AI Risk Management enhancements, DTEX has launched a suite of autonomous security agents aimed at streamlining investigations and threat analysis.
One of these tools, the Triage Guardian Agent, automates investigation workflows by collecting contextual evidence and applying structured oversight through independent reviewer agents. The system continuously analyses behavioural patterns before, during and after incidents, helping security teams understand how risks evolve while reducing false positives.
DTEX also introduced the Threat Hunter Agent, which supports proactive threat detection through agent-driven workflows. Security analysts can initiate investigations using natural language commands, allowing the platform to determine investigative steps, correlate findings and identify potential risks automatically. The solution incorporates more than 25 years of DTEX threat-hunting expertise, including collaboration with MITRE and Five Eyes defence partners.
According to DTEX, early deployments of its autonomous security capabilities have delivered productivity gains of more than 40 hours per month per analyst, enabling security teams to focus more on proactive risk management.
The enhanced AI Risk Management platform is currently available through a private preview programme, with broader market availability expected next quarter.
