As AI continues to transform how organizations operate, we are excited to announce Immersive AI Security. Our latest innovation tests the AI agents your teams build and governs how employees use AI tools, then retrains the people behind each risk through Immersive One.
Traditional security approaches weren't designed to monitor which AI tools employees are using or what data those tools is accessing. And while a growing number of AI security solutions can identify risks, detection alone isn't enough. Organizations also need to address those risks and strengthen the skills of the people responsible for using, building, and securing AI.
Welcome Immersive AI Security
Today, we're introducing Immersive AI Security, a new family of products designed to help organizations secure the AI their people use and build, while strengthening the human capabilities behind safer AI adoption.
The new offering includes two purpose-built products: AI Govern & Control, which helps organizations manage workforce AI usage, and AI Observe & Assess, which provides visibility and security testing for internally developed AI agents.
Already recognized as aLeader in the “Forrester Wave™ Cybersecurity Skills And Training Platforms, Q1 2026,” the new products further extend Immersive's proven approach to cyber readiness beyond the workforce to the AI technologies organizations increasingly depend on.
AI Govern & Control: scale AI while maintaining control
Employees aren't waiting for formal AI strategies to take shape. They're already using tools like ChatGPT, Claude, Co-Pilot, and Gemini to draft content, analyze information, write code, and complete everyday tasks.
But this widespread adoption creates challenges for security teams. Sensitive customer information, credentials, and proprietary business data can be inadvertently shared with AI tools, often without security teams knowing it's happening. Organizations shouldn't have to choose between embracing AI and protecting their data.
The solution provides visibility into workforce AI usage, including approved and unsanctioned tools, helping security leaders understand their organization's AI footprint and potential exposure.
When employees attempt to share sensitive information through browser-based AI tools such as Claude, ChatGPT or Gemini, Govern & Control detects the risk and delivers a contextual warning before the information is submitted. Employees can reconsider what they're sharing without unnecessarily disrupting their work.
Importantly, sensitive-data classification happens directly on the employee's device, preserving the privacy of their prompts and ensuring no raw prompt content leaves the browser.
But Govern & Control goes beyond identifying risky interactions. By connecting security findings to Immersive One, organizations can reinforce safer AI behavior through targeted upskilling informed by real-world activity. For example, an employee who pastes customer data into a chatbot gets a short lab that replays that moment. The security team then sees whether the same behavior happens again for that team.
The result is a more practical approach to AI governance that helps organizations manage risk while giving employees the confidence to use AI responsibly.
AI Observe & Assess: know and secure your agents before they go live
As organizations move from experimenting with generative AI to deploying autonomous agents, they're introducing a different set of security challenges.
Unlike traditional applications, AI agents can independently interact with tools, access data, connect to external services, and execute actions on behalf of users. Without sufficient visibility (such as mapping every tool, data source and MCP server an agent can reach) and testing, organizations may struggle to understand what these agents can reach, how they behave, or where vulnerabilities exist.
The solution combines agent observability with adversarial security testing, giving engineering and security teams visibility into agent connections, dependencies, and potential exposure.
Using testing aligned with established security frameworks, including the OWASP Top 10 for LLM Applications and MITRE ATLAS, teams can evaluate agents against known attack techniques and uncover vulnerabilities before attackers exploit them.
Findings can be incorporated into existing developer workflows such as Jira or Slack, helping teams prioritize remediation, retest fixes, and make more informed deployment decisions.
And through its connection to Immersive One, security insights can also inform developer and security team upskilling, helping strengthen the practices behind secure AI development.
Going beyond detection to build lasting AI security
What makes Immersive AI Security different is the ability to use one vendor to manage both employees and the AI they use and build.
Behind every AI interaction is a person deciding what information to share. Behind every autonomous agent is a team making decisions about its design, permissions, and security. Technology can identify vulnerabilities and risky behavior, but people ultimately play a critical role in addressing them.
That's why Immersive AI Security connects security findings with the exercising and upskilling capabilities of Immersive One. By bringing together visibility, assessment, actionable findings, and human readiness, we're helping organizations move beyond simply identifying AI security risks toward addressing the decisions and behaviors that create them.
AI adoption will continue to accelerate, and security strategies must evolve alongside it. With Immersive AI Security, we're expanding our mission to help organizations prove and improve their cyber readiness in an AI-driven world.
Because the question is no longer whether your organization is embracing AI; it's whether you can understand how employees use AI, test the agents you build for vulnerabilities, and prove your AI controls are working
If you’re interested in trying Immersive AI Security or want to learn more, please visit here.



