Context is the new perimeter

Midnight logins from odd places. Copy-paste into risky tools. Endpoints outside the office wall. The answer is not louder alarms, but smarter context: Al that learns behavior, adapts trust, and works with humans to keep work flowing and data safe

- PCQ Bureau [e-mail]

Picture an admin account logging in from an unusual location at 00:15, then pulling far more data than usual. In the old world, that might slip through because the user passed multi-factor authentication and the device looked healthy. In the new world, that is not enough. In a conversation with Sujatha Iyer, Head of AI Security, ManageEngine, a clear idea takes shape: security must learn in context, adjust in real time, and still leave room for human judgment. The journey starts with how Al was woven into the stack long before it became a buzzword, and lands on a simple principle: trust must be dynamic.

From intern to Al security lead:

Building with context

The arc begins in 2015 with an internship and grows into leading the Al for Security division. Along the way, the portfolio spans identity and access management, monitoring, mobile device management, and endpoint control. Across these products, contextual Al features are embedded. The stack evolves from Hadoop and big data to machine learning, neural networks, transformers, and now large language models. The throughline is constant: use context to make better, faster security decisions.

Trust that adapts:

Beyond MFA and posture

Static checks, multi-factor authentication (MFA) and device posture, set the entry rules, but they do not tell the whole story. User and entity behavior analytics (UEBA) adds the missing layer by learning patterns and assigning continuous risk. An admin logging in from an unusual location at midnight is flagged proactively. Adaptive MFA, paired with Al-driven risk scoring, turns trust into a living score that shifts as behavior shifts: dynamic and context-aware, not one-and-done.

Compliance by design:

Contextual intelligence