Magic Mirror now detects when users are logged into ChatGPT with personal accounts—or not logged in at all—and flags when their data may be used to train the model. We’re also adding customizable in-app messaging so you can guide users directly within their GenAI workflows.
Let’s dig into what’s new—and why it matters.
One of the biggest blind spots in GenAI security? Personal accounts.
Employees frequently toggle between personal and corporate accounts in tools like ChatGPT, Google Workspace, Microsoft 365, and Dropbox—all in the same browser session. This blurs the line between managed and unmanaged environments, introducing massive risk that traditional tools can’t catch.
With this release, Magic Mirror adds real-time Identity Assurance for ChatGPT sessions. In the Admin Session Audit dashboard, you’ll now see:
A 🔴 red indicator next to sessions when:
Hover over the dot to see which condition applies. This gives security and IT teams something they’ve never had before: real-time, session-level visibility into identity context inside GenAI tools.
And because Magic Mirror works on-device, this happens without sending user data to the cloud or relying on network logs that miss the nuance.
This is the first step in our broader Identity Assurance roadmap—support for additional domains is already in the works.
It’s not enough to write a GenAI policy and hope employees follow it.
With our new Custom Messages feature, you can now deliver short, targeted messages directly in the Magic Mirror User Console—the interface users see when they click Marv inside any GenAI tool.
You can use this to:
Security isn’t about stopping productivity—it’s about guiding it safely. This helps you do just that, without slowing down your team’s work.
This release isn’t just about risk reduction—it’s about insight.
MagicMirror helps organizations:
With every session observed, you're not just securing the organization—you’re also building a strategic view of how GenAI is transforming your workforce. And because it all happens on-device, you're gaining this insight without ever compromising user trust or privacy.
Personal accounts are a major—and often invisible—source of data loss. As employees blur the lines between work and personal tools, traditional security tools struggle to track where corporate data ends up. MagicMirror is the only GenAI security platform that tackles this issue at the root: the device, the session, and the moment it happens.
✅ Real-time visibility into ChatGPT identity context
✅ Red flags when data might be used for training
✅ Custom messaging to influence behavior at the source
✅ Strategic insights into GenAI adoption and usage
