DTEX Launches AI Risk Management to Secure AI Agents

DTEX has expanded its AI Risk Management offering, broadening its platform to help organizations secure the use of generative AI applications and autonomous AI agents. As GenAI tools, copilots, and AI agents gain greater access to enterprise data, systems, and business processes, many security solutions remain unable to distinguish or assess the intent behind actions taken by either people or AI. DTEX addresses this challenge with AI Risk Management, a comprehensive set of AI-native capabilities that uses behavioral intelligence to identify and mitigate risks originating from both human users and AI agents, delivering rapid and precise threat detection and response.

“AI agents are rapidly becoming operational actors inside the enterprise, with the ability to access data, interact with systems, and take autonomous action,” said Marshall Heilman, CEO of DTEX. “Many security solutions can monitor AI activity, but they still can’t determine whether that behavior aligns with intent or introduces risk. DTEX AI Risk Management solves that by applying behavioral intelligence to understand and respond to risk across both humans and AI agents, enabling organizations to operate securely at the speed and scale modern AI environments demand.”

By combining AI risk management with autonomous investigation and response, DTEX enables organizations to accelerate AI adoption with the visibility, control, and operational confidence required to safely scale AI-driven productivity and innovation across the enterprise.

Monitor and protect AI activity

As AI agents begin operating autonomously across enterprise systems, organizations face a new category of risk. Unlike traditional software, AI agents can interpret instructions, access sensitive data, interact with external systems, and make decisions with limited human oversight. Securing these environments requires more than activity monitoring. It requires understanding what the agent was instructed to do, how behavior evolves over time, and whether actions align with expected intent.

DTEX delivers comprehensive visibility into how AI is used across the enterprise and applies deep behavioral context to identify emerging risk before it becomes a breach.

With AI Risk Management organizations can:

Discover sanctioned and unsanctioned AI usage across users, endpoints, and workflows, including browser, IDE, application, and embedded AI activity.
Identify shadow AI and embedded copilots in real time, dynamically building sanctioned tool inventories and automatically classifying the risk of unknown or unmanaged AI tools.
Monitor prompts, responses, and data movement at a granular level, including uploads, downloads, and AI-generated content, to detect leakage of source code, intellectual property, and sensitive enterprise data.
Classify prompts and interactions to support auditing, compliance, and threat investigations, enabling security teams to understand not just what was asked, but why, through behavioral context and intent analysis.
Analyze AI activity to infer both human and AI agent intent, distinguishing normal experimentation from risky or malicious behavior by correlating prompts, historical patterns, behavioral baselines, and agent actions over time.
Differentiate human versus AI-driven actions and deliver deep visibility into “Computer Use” AI (CUI), including what an agent was instructed to do, how it executed tasks, and the detailed lineage of actions performed across enterprise systems.
Detect and prevent autonomous agent-driven data exfiltration using behavioral monitoring, prompt lineage, and AI risk models that proactively identify high-risk agentic behavior and the intersection between human and AI risk

In one early deployment, DTEX identified an autonomous AI agent exposing sensitive enterprise data despite operating within its intended workflow and permissions. By correlating prompt lineage, behavioral patterns, and contextual activity over time, DTEX surfaced the risk before it resulted in a security incident.

Act on risk with autonomous security agents

To make AI Risk Management operational, DTEX is also introducing autonomous security agents that apply behavioral context and risk modeling to automate investigation and threat analysis. This enables organizations to differentiate human vs AI-driven activity, track behavioral patterns over time, and understand how AI systems interact with data and identities.

Triage Guardian Agent

Built on more than 20 years of DTEX i³ behavioral expertise, Triage Guardian applies a multi-agent approach to deliver consistent, defensible triage outcomes at scale. Unlike traditional alert-driven workflows that evaluate isolated events, Triage Guardian continuously analyzes behavioral context before, during, and after an incident, allowing agents to effectively rewind and fast-forward investigative timelines to understand how risk evolved over time. It automates investigation workflows, gathers contextual evidence, and applies structured human oversight through independent reviewer agents that validate findings, minimize bias, and ensure conclusions remain evidence-backed. By combining behavioral intelligence with analyst-grade decision logic, Triage Guardian dramatically reduces false positives while minimizing missed risks that conventional triage approaches often fail to detect.

Threat Hunter Agent

Threat Hunter enables proactive threat discovery through agentic workflows, continuously assessing the evolving risk landscape, generating detailed threat analysis, and identifying previously unknown threats before they surface in an incident. Analysts can initiate complex threat hunts using natural language, allowing Threat Hunter to determine how to execute the investigation, correlate findings, and surface relevant risk autonomously. Built on more than 25 years of DTEX i³ threat hunting expertise, including collaborative research with MITRE and FVEY defense partners, Threat Hunter applies proven analyst tradecraft and investigative context to every hunt at machine speed.

Early deployments demonstrate more than 40 hours saved per month per analyst.

“Within weeks of deploying DTEX Autonomous Triage, our analysts saw immediate value. What would normally take hours to investigate is now surfaced instantly with clear context and explanations. It has improved the quality of our investigations, reduced fatigue, and allowed us to focus on proactive security work.” — Assistant Director, Cyber Operations, Government Agency