Incidents and Risk Signals – Overview

Prev Next

Modern social engineering has evolved well beyond traditional phishing emails. Today’s attacks increasingly rely on human-in-the-loop techniques, including direct interaction with help desks, executive impersonation using AI-generated voices, and the use of remote or intermediary systems to appear geographically and operationally legitimate.

These attacks routinely bypass traditional perimeter defenses such as firewalls, web application firewalls (WAFs), and IP-based controls. In many cases, attackers operate with valid credentials, access systems through real browsers, and deliberately mimic legitimate user behavior. As a result, network-centric and credential-centric security controls alone are insufficient.

The detectors described in this section are designed to identify the technical seams and inconsistencies that emerge when an attacker attempts to impersonate a legitimate user. Rather than focusing on a single signal, these detectors correlate browser behavior, device characteristics, execution environment, network routing, and organizational context. By cross-referencing real-time session data with authoritative sources such as identity provider history, HR records, and known user baselines, the system surfaces risk indicators that are difficult for attackers to fully replicate - even when authentication succeeds.


Detector Framework

The detection framework is organized into logical categories that reflect the different layers an attacker must control in order to convincingly impersonate a real user. Each detector contributes contextual risk signals, not binary verdicts, and is intended to be evaluated in combination with others.

Signals are collected at three levels of depth, depending on how Imper.ai is deployed:

  • Web-based detection 0 Active JavaScript running in the browser at the moment of a session or interaction, requiring no software installation on the user's device.

  • Point-in-time device scan - A lightweight script run on managed endpoints at the time of a verification event, surfacing device-level context not visible from the browser alone.

  • Deployed agent - A continuously running agent on managed devices that builds behavioral baselines over time and detects persistent or evolving risk signals.


Network and Location

Network and location signals evaluate where access appears to originate and how traffic is routed to the platform. These detectors identify inconsistencies between reported location, routing infrastructure, and expected user geography.

Because social engineering attacks often rely on geographic plausibility to support impersonation pretexts, anomalies in network path or location frequently expose attempts to obscure true origin.

Detector

Primary Social Engineering Risk

Masked Network Connection

Use of Tor, proxies, or anonymization infrastructure to hide origin

VPN

Impersonation of remote workers or geographic alignment with a victim

Anomalous Location

Location mismatches relative to HR records, CVs, or identity provider history

Laptop Farm

Multiple employee identities sharing a residential IP or subnet, indicating a device farm or coordinated fraud operation rather than independent remote workers

Endpoint

Endpoint signals focus on the device and execution environment used to access the platform. These detectors evaluate whether the hardware, operating system, browser, media capabilities, and control layer align with what is expected for the claimed user.

Attackers frequently rely on disposable, virtualized, remotely operated, or specially configured environments. While credentials can be stolen, reproducing a legitimate endpoint profile over time is significantly more difficult.

Detector

Primary Social Engineering Risk

Virtual Machine

Use of disposable or isolated attack environments

Virtual Audio Devices

Injection of AI-generated or pre-recorded audio into voice interactions

Device Mismatch

Session or identity hijacking from an unauthorized machine

Anomalous Device

Use of attack rigs or low-end hardware inconsistent with the user profile

Anomalous Browser

Use of anti-detect or identity-spoofing browsers

Remote Access & Control

Hardware (KVM-over-IP) or software remote desktop tools used to operate the device from a hidden third-party location

Behavior and Usage

Behavior and usage signals analyze how a user acts during a session and over time on their device. These detectors span a range of visibility: some signals are captured from the browser alone, while others, particularly those involving usage habits or input patterns, require a point-in-time scan or the deployed agent.

In social engineering campaigns, behavioral anomalies emerge when attackers use automation, assume another person's identity, or operate a device that is not their own.

Detector

Primary Social Engineering Risk

Bot-like Behavior

Automation of credential abuse, brute-force attempts, or MFA guessing

Inconsistent Patterns

Deviations from established technical fingerprints and usage habits: browser, OS, working hours, application patterns

Suspected Attack Device

Presence of browser-visible attacker tooling: debuggers, hacking extensions, or automation artifacts

Clean / Burner Device

Use of freshly created or wiped environments with no usage history

Malicious Tooling on Device

Presence of OS-level hacking tools, exploit scripts, or attacker-oriented repositories — detected by scan or agent

Anomalous Work-Only Profile

Device with months of history but zero personal activity, indicating a purpose-built impostor environment

Anomalous Input Patterns

Typing timing, key hold duration, and mouse movement shifts indicating a different person or automated input is operating the device